Molecular identification and expression of erythroid K:Cl cotransporter in human and mouse erythroleukemic cells.

A major pathway for K+ efflux in human reticulocytes and young RBCs is K:Cl cotransport (K:Cl-CT). The activity of K:Cl-CT is increased in pathologic RBCs containing hemoglobins S and C and may contribute to the abnormal dehydration state of these cells. Human K:Cl-CT (gene product KCC1) has been recently sequenced from human (hKCC1), rabbit and rat tissue by Gillen et al. (J Biol Chem 271:16237, 1996). We report here the sequence of KCC1 from human and mouse erythroleukemic cells (K562 and MEL cells, respectively). The cDNA for human erythroid-KCC1 is 100% identical to hKCC1 and the cDNA for mouse erythroid-KCC1 shares 89% identity with hKCC1, which translates to 96% identity at the amino acid level. Mammalian KCC1 is strongly conserved with >95% identity between human, rabbit, rat, and mouse KCC1 proteins. We did not detect any full-length mRNA transcripts of human erythroid-KCC1 in circulating reticulocytes. We detected two mRNA isoforms of human erythroid-KCC1 that resulted in C-terminal truncated proteins (73 amino acid and 17 amino acids, respectively). Human and mouse erythroidKCC1 differed at several consensus sites including a predicted PKC phosphorylation site at 108threonine and a predicted CK2 phosphorylation site at 51serine, within the predicted cytoplasmic N-terminal, that are present in human but not mouse erythroid-KCC1. Expression of MEL-KCC1 mRNA increases substantially upon DMSO-induced differentiation opening the possibility that erythroid-KCC1 plays a role in early erythroid maturation events. The molecular identification of erythroid-KCC1 is an important step towards understanding the physiologic role mediated by this protein in young and pathologic RBCs and during erythropoiesis, as well as providing a new tool for the elucidation of pathways and signals involved in RBC volume regulation.

Molecular cloning and expression of a chloride channel-associated protein pICln in human young red blood cells: association with actin.

We report the cloning and sequencing from human reticulocytes of cDNA coding for the Cl- channel-associated protein, pICln. Human reticulocyte pICln (HRpICln) cDNA encodes a protein (predicted molecular mass 26293Da) identical with human non-pigmented ciliary epithelial cell pICln. By using full-length HRpICln cDNA (approx. 1.2 kb) to probe human lymphocyte metaphase-chromosome spreads, the location of the human ICln gene was mapped to 11q13 by fluorescence in situ hybridization analysis. Polyclonal antibodies to recombinant HRpICln detected bands at approx. 43 kDa and approx. 37 kDa in both normal (AA) and sickle (SS) red blood cell (RBC) ghost membranes. In SS ghosts, and in ghosts from a patient with autoimmune haemolytic anaemia with 9.8% reticulocytes, the amount of HRpICln was increased compared with AA ghosts, suggesting that the expression or membrane assembly of HRpICln is cell age-dependent. Laser scanning confocal fluorescent microscopy immunolocalized HRpICln largely to the RBC membrane. The increased staining intensity of HRpICln in a reticulocyte-enriched AA RBC density-separated fraction is consistent with a dependence of HRpICln membrane content on cell age. HRpICln and beta-actin form stable complexes in vivo, demonstrated with the yeast two-hybrid system. Low-ionic-strength extraction of ghost membranes, which results in the extraction of the spectrin-actin cytoskeleton, also results in the extraction of HRpICln, consistent with the possibility for the association of these proteins in RBCs in vivo. The results presented here establish the presence of the Cl- channel-associated protein, pICln, in human RBCs, and raises the possibility that this protein has a role in RBC Cl- transport and volume regulation in young RBCs. Moreover the association of RBC pICln with actin offers a model in which to test interactions between RBC ion channels and the cytoskeleton.

A potential determinant of enhanced crystallization of Hbc: spectroscopic and functional evidence of an alteration in the central cavity of oxyHbC.

The structural basis of the crystallizing tendencies of oxyHbC (beta 6Glu-->Lys), that produces haemolytic anaemia in homozygotes, is unknown. Using a fluorescent organic phosphate analogue (8-hydroxy-1,3,6-pyrenetrisulphonate), and conventional oxygen equilibrium studies, data suggest that the binding of inositolhexaphosphate (IHP) to oxyHbC differs from HbA, indicating perturbations of the oxyHbC central cavity, which was predicted from our earlier spectroscopic findings. To define the relationship between this conformational change in oxyHbC and its tendency to crystallize, the effect of four central cavity ligands on the crystallization rate was studied: a peptide containing 11 residues from the N-terminal portion of band 3, the full cytoplasmic domain of band 3, 2,3-diphosphoglycerate and IHP. OxyHbC crystallization was accelerated by all these central cavity ligands and not by the appropriate controls. These central cavity changes become an excellent candidate for the dramatic increase in the crystallization rate of oxyHbC.

Increased rotational mobility and extractability of band 3 from protein 4.2-deficient erythrocyte membranes: evidence of a role for protein 4.2 in strengthening the band 3-cytoskeleton linkage.

Band 3 (anion-exchange protein 1-[AE1]) is the major integral membrane protein of human erythrocytes and links the membrane to the underlying cytoskeleton via high-affinity binding to ankyrin. It is unclear whether other cytoskeletal proteins participate in strengthening the ankyrin-band 3 linkage, but a putative role for protein 4.2 (P4.2) has been proposed based on the increased osmotic fragility and spherocytic morphology of P4.2-deficient red blood cells (RBCs). The present study was designed to investigate the hypothesis that P4.2 has a direct role in strengthening the band 3-cytoskeleton linkage in human RBCs, by measuring independent features of this interaction in normal and P4.2-deficient RBCs. The features examined were the rotational mobility of band 3 assayed by time-resolved phosphorescence emission anisotropy (TPA), and the extractability of band 3 by octyl-beta-glucoside, the latter being a nonionic detergent that selectively extracts only band 3 that is not anchored to the cytoskeleton. We find that the amplitude of the most rapidly rotating population of band 3 (correlation time, approximately 30 to 60 microseconds) is increased 81% and 67% in P4.2-deficient ghosts (P4.2NIPPON and band 3MONTEFIORE, respectively) compared with control ghosts. The amplitude of the intermediate speed rotating population of band 3 (correlation time, approximately 200 to 500 microseconds) is increased 23% and 8% in P4.2-deficient ghosts (P4.2NIPPON and band 3MONTEFIORE, respectively) compared with control ghosts, at the expense of the slowly rotating component (correlation time, approximately 2,000 to 3,000 microseconds, amplitude decreased 43% and 39% in P4.2NIPPON and band 3MONTEFIORE, respectively) and immobile component (immobile on this experimental time scale; amplitude decreased 26% and 10% in P4.2NIPPON and band 3MONTEFIORE, respectively) of band 3. These results demonstrate that P4.2 deficiency partially removes band 3 rotational constraints, ie, it increases band 3 rotational mobility. The nonionic detergent octyl-beta-glucoside, which does not disturb band 3-cytoskeleton associations, ie, it extracts only band 3 that is not attached to the cytoskeleton, extracted 30% and 61% more band 3 from P4.2NIPPON and band 3MONTEFIORE ghost membranes, respectively, compared with control ghosts. The octyl-beta-glucoside ghost extracts from both P4.2-deficient phenotypes were enriched in band 3 oligomeric species (tetramers, higher-order oligomers, and aggregates) compared with controls. Since band 3 oligomers selectively associate with the cytoskeleton, these results are consistent with a weakened band 3-cytoskeleton linkage in P4.2-deficient RBC membranes. P4.2 deficiency does not affect band 3 anion transport activity, since uptake of radiolabeled sulfate was similar for control and P4.2-deficient RBCs. Thus, we propose that P4.2 directly participates in strengthening the band 3-cytoskeleton linkage.

Decreased content of protein 4.2 in ankyrin-deficient normoblastosis (nb/nb) mouse red blood cells: evidence for ankyrin enhancement of protein 4.2 membrane binding.

Homozygous normoblastosis (nb/nb) mice, whose red blood cell (RBC) membranes are nearly completely deficient in full-length 210-kD ankyrin, were used to study interactions between ankyrin and protein 4.2 (P4.2). Although it is unclear whether or not these proteins interact in the membrane, both ankyrin and P4.2 bind to the cytoplasmic domain of band 3 (cdb3). In addition to the complete deficiency of full-length ankyrin, nb/nb RBC membranes are also partially spectrin deficient, resulting in morphologically spherocytic and mechanically fragile cells. A new finding was that nb/nb RBC membranes are severely (approximately 73%) P4.2 deficient compared with wild type (+/+) or high reticulocyte mouse RBC membranes. Metabolic labeling of nb/nb reticulocytes showed active P4.2 synthesis at levels comparable with high reticulocyte controls suggesting that the nb/nb P4.2 deficiency was not the result of defective P4.2 synthesis. Reconstitution of nb/nb inside-out vesicles (IOVs) with human RBC ankyrin restored ankyrin levels to approximately 80% of +/+ IOV levels and increased binding of exogenously added human RBC P4.2 by approximately 60%. These results suggest that ankyrin is required for normal associations of P4.2 with the RBC membrane.

Identification of a band-3 binding site near the N-terminus of erythrocyte membrane protein 4.2.

Protein 4.2 (P4.2) is a major component of the erythrocyte plasma membrane accounting for approx. 5% of total membrane protein. The major membrane binding site for P4.2 is contained within the cytoplasmic domain of band 3 (cdb3), although the precise location of the cdb3 binding site is not known. To identify the cdb3 binding site, we used synthetic P4.2 peptides (15-mers) that spanned the entire 721-amino-acid large isoform of P4.2, and determined the binding of these peptides to cdb3 in an in vitro binding assay. One peptide, P8 (L61FVRRGQPFTIILYF), bound strongly to cdb3 and four others bound less strongly (P22, L271LNKRRGSVPILRQW; P27, G346EGQRGRIWIFQTST; P41, L556WRKKLHLTLSANLE; P48, I661HRERSYRFRSVWPE). These peptides have in common a cluster of two or three basic amino acid residues (arginine or lysine), in a region without nearby acidic residues. Cdb3 bound saturably to P8 with a Kd of 0.16 microM and a capacity of 0.56 mol of cdb3 monomer/mol of P8. Use of overlapping synthetic peptides further defined the cdb3 site as being contained within V63RRGQPFTIILYF. Replacement of R64R with R64G, G64R or G64G almost completely abolished cdb3 binding, suggesting that R64R is essential for cdb3 binding. P8 competitively inhibited binding of purified human erythrocyte P4.2 to cdb3. In blot overlay assays, cdb3 bound to a 23 kDa N-terminal P4.2 tryptic peptide containing V63RRGQPFTIILYF but not to other P4.2 tryptic peptides lacking this site. The V63RRGQPFTIILYF site is highly conserved in mouse and human erythrocyte P4.2 as well as between P4.2 and transglutaminase proteins, which are evolutionarily related to P4.2.

Molecular cloning of mouse erythrocyte protein 4.2: a membrane protein with strong homology with the transglutaminase supergene family.

We report the molecular cloning and characterization of mouse erythrocyte protein 4.2 (P4.2). Mouse erythrocyte P4.2 is a 691-amino-acid protein with a predicted MW of 77 kDa. Northern blot analysis detected a 2.2-kb transcript in mouse reticulocytes, compared with a 2.4- to 2.5-kb transcript in human reticulocytes, which is consistent with the absence of the 30-amino-acid splicing insert in mouse erythrocyte P4.2 that is found in the human protein (isoform I). Like the human erythrocyte P4.2, mouse erythrocyte P4.2 contains regions strikingly homologous with the transglutaminase (TGase) proteins although it too most likely lacks TGase crosslinking activity. Mouse P4.2 is on average 73% identical with human erythrocyte P4.2, although regional variations exist, with greatest conservation in the regions of the molecule that contain the TGase active site, the TGase calcium-binding site, and a band 3 binding site. Hydropathy analysis reveals a protein containing a series of hydrophobic domains, similar to the situation for human P4.2 and consistent with its tight binding to the membrane, although the mouse P4.2 is missing both the strongly hydrophilic region and adjacent highly charged region that are present in the human protein, suggesting that the two proteins could differ in their physical characteristics, binding associations, or functional properties. The availability of the complete mouse erythrocyte P4.2 cDNA should help in the design of P4.2-deficient animal models (for example, ribozyme or homologous recombinant "knockout" models) that should accelerate the understanding of P4.2 function in both erythroid and non-erythroid cells.

Human erythrocyte protein 4.2 deficiency associated with hemolytic anemia and a homozygous 40glutamic acid-->lysine substitution in the cytoplasmic domain of band 3 (band 3Montefiore).

Red blood cell (RBC) protein 4.2 deficiency is often associated with a moderate nonimmune hemolytic anemia, splenomegaly, and osmotically fragile RBCs resembling, but not identical to, hereditary spherocytosis (HS). In the Japanese type of protein 4.2 deficiency (protein 4.2Nippon), the anemia is associated with a point mutation in the protein 4.2 cDNA. In this report, we describe a patient with moderate and apparently episodic nonimmune hemolytic anemia with splenomegaly, spherocytosis, osmotically fragile RBCs, reduced whole cell deformability, and abnormally dense cells. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of the proposita's RBC membrane proteins showed an 88% deficiency of protein 4.2 and a 30% deficiency of glyceraldehyde-3-phosphate dehydrogenase (band 6). Structural and molecular analyses of the proposita's protein 4.2 were normal. In contrast, limited tryptic digestion of the proposita's band 3 showed a homozygous abnormality in the cytoplasmic domain. Analysis of the pedigree disclosed six members who were heterozygotes for the band 3 structural abnormality and one member who was a normal homozygote. Direct sequence analysis of the abnormal band 3 tryptic peptide suggested that the structural abnormality resided at or near residue 40. Sequence analysis of the proposita's band 3 cDNA showed a 232G-->A mutation resulting in a 40glutamic acid-->lysine substitution (band 3Montefiore). Allele-specific oligonucleotide hybridization was used to probe for the mutation in the pedigree, showing that the proposita was homozygous, and the pedigree members who were heterozygous for the band 3 structural abnormality were also heterozygous for the band 3Montefiore mutation. The band 3Montefiore mutation was absent in 26 chromosomes from race-matched controls and in one pedigree member who did not express the band 3 structural abnormality. In coincidence with splenectomy, the proposita's anemia was largely corrected along with the disappearance of most spherocytes and considerable improvements of RBC osmotic fragility, whole cell deformability, and cell density. We conclude that this hereditary hemolytic anemia is associated with the homozygous state for band 3Montefiore (40glutamic acid-->lysine) and a decreased RBC membrane content of protein 4.2. We speculate that band 3 structural abnormalities can result in defective interactions with protein 4.2 and band 6, and in particular, that the region of band 3 containing 40glutamic acid is involved directly or indirectly in interactions with these proteins.

Characterization of the phosphorylated state of protein 4.2 from a patient partially deficient in protein 4.2.

These studies compare the protein 4.2 found in a patient with osmotically fragile, spherocytic erythrocytes to the normal protein 4.2. The patient protein 4.2 is present in the erythrocyte ghost membranes as a doublet of 74 and 72 KDa at a concentration less than 1% of normal. The patient protein 4.2 becomes highly phosphorylated in the presence of Zn++ and is phosphorylated, relative to the amount of protein present, to a greater extent than the normal 72 KDa protein 4.2. These studies indicate that both the patient and the normal protein 4.2 usually exists in a highly phosphorylated state. The phosphorylation sites on the patient protein 4.2 appear to be more readily cycled than on the normal protein 4.2. Staphylococcus aureus V8 protease generates similar phosphopeptides in both the normal and patient protein 4.2 except for an extra 11 KDa phosphopeptide generated from the 74 KDa form of the protein.

An alanine-to-threonine substitution in protein 4.2 cDNA is associated with a Japanese form of hereditary hemolytic anemia (protein 4.2NIPPON).

Erythrocyte (RBC) protein 4.2 (P4.2)-deficiency observed in Japanese individuals results in a hemolytic anemia associated with abnormally shaped (spherocytic, ovalocytic, and elliptocytic), osmotically fragile RBCs, the clinical presentation of which resembles hereditary spherocytosis (HS). By sodium dodecyl sulfate-polyacrylamide gel electrophoresis, P4.2-deficient individuals contain less than 1% of the normal membrane content of P4.2 and immunologic analysis shows that the P4.2 present exists as an equimolar doublet of 74-Kd and 72-Kd bands, in contrast to normal RBC membranes where a discrete 74-Kd band is not observed. RBC membranes from both of the biologic parents of a P4.2-deficient individual contained both the 74-Kd and the 72-Kd bands, demonstrating their heterozygosity for the P4.2 defect. The molecular basis of Japanese P4.2-deficiency was investigated by reverse transcription of total reticulocyte RNA, followed by polymerase chain reaction (PCR) amplification, subcloning, and sequencing. The complete cDNA sequence of a P4.2-deficient patient showed a single point mutation that changes codon 142 from GCT (alanine) to ACT (threonine) (Protein 4.2NIPPON). The mutation also eliminated an HgaI restriction site, therefore allowing rapid screening for the presence of the mutation. Screening of PCR-amplified genomic DNA showed that the mutation was present in the homozygous state in four (eight chromosomes) unrelated Japanese P4.2-deficient individuals and absent in 35 (70 chromosomes) P4.2-normal controls (including 15 Japanese [30 chromosomes]). The presence of the mutation was confirmed by allele-specific hybridization. The mutation occurred in an alternatively spliced exon that is present in two of four P4.2 mRNA splicing isoforms. These results demonstrate that Japanese P4.2-deficiency is closely associated with the P4.2 gene and does not arise secondarily to a defect in another membrane protein, and further suggest that the P4.2-deficiency is related to the pathogenesis of the hemolytic anemia in this variant form of recessively inherited spherocytosis.

The gene for human erythrocyte protein 4.2 maps to chromosome 15q15.

Protein 4.2 (P4.2), one of the major components of the red-blood-cell membrane, is located on the interior surface, where it binds with high affinity to the cytoplasmic domain of band 3. Individuals whose red blood cells are deficient in P4.2 have osmotically fragile, abnormally shaped cells and moderate hemolytic anemia. cDNA clones from both the 5' and the 3' coding regions of the P4.2 gene were used to map its chromosomal location by fluorescence in situ hybridization. The probes, individually or in combination, gave specific hybridization signals on chromosome 15. The hybridization locus was identified by combining fluorescence images of the probe signals with fluorescence banding patterns generated by Alu-PCR (R-like) probe and by DAPI staining (G-like). Our results demonstrate that the locus of the P4.2 gene is located within 15q15.

Oxidation of spectrin and deformability defects in diabetic erythrocytes.

We reasoned that de novo oxidative damage, as a result of increased protein glycosylation, could participate in the mechanisms whereby diabetic erythrocytes acquire membrane abnormalities. To examine this hypothesis, the extent of erythrocyte membrane protein glycosylation and the oxidative status of spectrin, the major component of the erythrocyte membrane skeleton, were examined. Labeling erythrocyte membranes with [3H]borohydride, which labels glucose residues bound to proteins, revealed that several proteins were heavily glycosylated compared with nondiabetic erythrocyte membranes. In particular, the proteins beta-spectrin, ankyrin, and protein 4.2 were the most glycosylated. Although sodium dodecyl sulfate-polyacrylamide gel electrophoresis of diabetic erythrocyte membranes did not reveal any quantitative or qualitative abnormalities in spectrin or other membrane proteins, examination of spectrin oxidative status by amino acid analysis and with cis-dichlorodiammineplatinum(II) (cDDP), a chemical probe specific for protein methionine and cysteine residues, demonstrated that the diabetic spectrin was oxidatively damaged: spectrin from diabetic subjects contained 35% less methionine (P less than 0.002), 15% less histidine (P less than 0.006), and a twofold increase in cysteic acid (P less than 0.001) compared with normal spectrin. Diabetic spectrin bound 32% less cDDP than normal spectrin (P less than 0.001); the lowest cDDP binding was observed with spectrin from insulin-dependent diabetic subjects. The extent of cDDP binding to diabetic spectrin correlated moderately and inversely with glycosylated hemoglobin (GHb) levels (n = 12, r = -0.727). Erythrocyte deformability, measured by ektacytometry, was decreased between 5 and 23% of control measurements (average of approximately 10%) in 21 of 32 diabetic subjects surveyed.(ABSTRACT TRUNCATED AT 250 WORDS)

Growth of Plasmodium falciparum in human erythrocytes containing abnormal membrane proteins.

To evaluate the role of erythrocyte (RBC) membrane proteins in the invasion and maturation of Plasmodium falciparum, we have studied, in culture, abnormal RBCs containing quantitative or qualitative membrane protein defects. These defects included hereditary spherocytosis (HS) due to decreases in the content of spectrin [HS(Sp+)], hereditary elliptocytosis (HE) due to protein 4.1 deficiency [HE(4.1(0))], HE due to a spectrin alpha I domain structural variant that results in increased content of spectrin dimers [HE(Sp alpha I/65)], and band 3 structural variants. Parasite invasion, measured by the initial uptake of [3H]hypoxanthine 18 hr after inoculation with merozoites, was normal in all of the pathologic RBCs. In contrast, RBCs from six HS(Sp+) subjects showed marked growth inhibition that became apparent after the first or second growth cycle. Preincubation of HS(Sp+) RBCs in culture for 3 days did not alter these results. Normal parasite growth was observed in RBCs from one HS subject with normal membrane spectrin content. The extent of decreased parasite growth in HS(Sp+) RBCs closely correlated with the extent of RBC spectrin deficiency (r = 0.90). Homogeneous subpopulations of dense HS RBCs exhibited decreased parasite growth to the same extent as did HS whole blood. RBCs from four HE subjects showed marked parasite growth inhibition, the extent of which correlated with the content of spectrin dimers (r = 0.94). RBCs from two unrelated subjects with structural variants of band 3 sustained normal parasite growth. Decreased growth in the pathologic RBCs was not the result of decreased ATP or glutathione levels or of increased RBC hemolysis. We conclude that abnormal parasite growth in these RBCs is not the consequence of metabolic or secondary defects. Instead, we suggest that a functionally and structurally normal host membrane is indispensable for parasite growth and development.

Molecular cloning of human protein 4.2: a major component of the erythrocyte membrane.

Protein 4.2 (P4.2) comprises approximately 5% of the protein mass of human erythrocyte (RBC) membranes. Anemia occurs in patients with RBCs deficient in P4.2, suggesting a role for this protein in maintaining RBC stability and integrity. We now report the molecular cloning and characterization of human RBC P4.2 cDNAs. By immunoscreening a human reticulocyte cDNA library and by using the polymerase chain reaction, two cDNA sequences of 2.4 and 2.5 kilobases (kb) were obtained. These cDNAs differ only by a 90-base-pair insert in the longer isoform located three codons downstream from the putative initiation site. The 2.4- and 2.5-kb cDNAs predict proteins of approximately 77 and approximately 80 kDa, respectively, and the authenticity was confirmed by sequence identity with 46 amino acids of three cyanogen bromide-cleaved peptides of P4.2. Northern blot analysis detected a major 2.4-kb RNA species in reticulocytes. Isolation of two P4.2 cDNAs implies existence of specific regulation of P4.2 expression in human RBCs. Human RBC P4.2 has significant homology with human factor XIII subunit a and guinea pig liver transglutaminase. Sequence alignment of P4.2 with these two transglutaminases, however, revealed that P4.2 lacks the critical cysteine residue required for the enzymatic crosslinking of substrates.

Deficiency of protein 4.2 in erythrocytes from a patient with a Coombs negative hemolytic anemia. Evidence for a role of protein 4.2 in stabilizing ankyrin on the membrane.

A patient with a mild hemolytic anemia and osmotically fragile, spherocytic erythrocytes was studied. Analysis of the erythrocyte membrane proteins by SDS-PAGE revealed a deficiency of protein 4.2 (less than 0.10% of normal). The protein 4.2-deficient erythrocytes contained normal amounts of all other membrane proteins, although the amount of band 3 was slightly reduced and the amount of band 6 (G3PD) was slightly elevated. The spectrin content of these cells was normal, as measured by both SDS-PAGE and radioimmunoassay. Erythrocytes from the patient's biologic parents were hematologically normal and contained normal amounts of protein 4.2. Immunological analysis using affinity purified antibodies revealed that the patient's protein 4.2 was composed of equal amounts of a 74-kD and 72-kD protein doublet, whereas the normal protein was composed primarily of a 72-kD monomer. Proteolytic digestion studies using trypsin, alpha-chymotrypsin and papain demonstrated that the patient's protein 4.2 was similar but not identical to the normal protein. Binding studies showed that the protein 4.2-deficient membranes bound purified protein 4.2 to the same extent as normal membranes, suggesting that the membrane binding site(s) for the protein were normal. Depleting the protein 4.2-deficient membranes of spectrin and actin resulted in a loss of nearly two-thirds of the membrane ankyrin, whereas similar depletion of normal membranes resulted in no loss of ankyrin. Repletion of the protein 4.2-deficient membranes with purified protein 4.2 before spectrin-actin extraction partially prevented the loss of ankyrin. These results suggest that protein 4.2 may function to stabilize ankyrin on the erythrocyte membrane.

Human erythrocyte protein 4.1 is a phosphatidylserine binding protein.

The aminophospholipids phosphatidylethanolamine (PE) and phosphatidylserine (PS) are the major phospholipids contained in the cytoplasmic leaflet of the human erythrocyte (RBC) plasma membrane and are largely confined to that leaflet over the entire RBC lifespan. In particular, PS, which comprises approximately 13% of total RBC membrane phospholipids, is normally restricted entirely to the cytoplasmic leaflet. However, molecular mechanisms that regulate this asymmetric distribution of phospholipids are largely unknown. We examined elliptocytic RBCs that completely lacked protein 4.1 (HE [4.1 degrees]), but contained normal amounts of all other peripheral membrane proteins, and found approximately 10% of total membrane PS was accessible in the exoplasmic leaflet of these membranes. Inside out vesicles (IOVs) derived from HE [4.1 degrees] RBCs bound fewer PS liposomes than did IOVs derived from normal RBCs. Normal IOVs that were depleted of proteins 2.1 (ankyrin), 4.1, and 4.2 bound fewer PS liposomes similar to HE [4.1 degrees] IOVs, and repletion with protein 4.1 restored PS liposome binding to control levels. Addition of purified protein 4.1 to PS liposomes resulted in saturable binding with the extent of binding being proportional to the liposome PS content. Our data suggests that human RBC protein 4.1 is a PS binding protein and may be involved in the molecular mechanisms that stabilize PS in the cytoplasmic leaflet of the human RBC plasma membrane.

Protein 4.1 in sickle erythrocytes. Evidence for oxidative damage.

Sickle erythrocytes are known to undergo excessive auto-oxidation, resulting in the generation of increased intracellular levels of several species of free radical oxidants. This environment is likely to enhance the accumulation of oxidative lesions by membrane components, although, as yet, this has been shown directly only for the sickle membrane phospholipids. We examined the oxidative status of protein 4.1, a major component of the human erythrocyte protein skeleton. We found that protein 4.1 isolated from sickle erythrocytes bound approximately 4-fold less to protein 4.1-stripped membranes than did the normal protein. The binding defect was inherent in the sickle protein and not in its membrane-binding site(s) since normal protein 4.1 bound to sickle protein 4.1-stripped inside-out vesicles similar to normal protein 4.1-stripped inside-out vesicles. Sickle membranes, in particular spectrin-depleted inside-out vesicles, contained less protein 4.1 than normal membranes. Purified sickle protein 4.1 contained 20-40% high molecular weight aggregated protein (Mr greater than 200,000), whereas the purified normal protein contained approximately 10% high molecular weight protein. The high molecular weight protein was immunoreactive with antibodies to protein 4.1 but not with antibodies to spectrin, ankyrin, band 3, glycophorin, or hemoglobin, suggesting that the high molecular weight protein was cross-linked protein 4.1 and not a complex of protein 4.1 and some other membrane protein(s). Purified sickle protein 4.1 was eluted from an anion-exchange resin at a higher salt concentration than normal protein 4.1. Oxidizing normal protein 4.1 with diamide resulted in an anion-exchange elution pattern similar to the sickle protein, suggesting that oxidation can affect protein surface charge. Activated thiol beads bound one-half as much sickle protein 4.1 as normal protein 4.1 when both were solubilized directly from membranes, demonstrating that thiol oxidation had occurred in vivo. Direct quantification of protein thiols revealed that the sickle protein contained 1-2 mol% fewer cysteines/protein 4.1 monomer than did the normal protein. By amino acid analysis, sickle protein 4.1 was found to contain less methionine and tyrosine than did the normal protein and contained approximately 1 mol% cysteic acid, whereas the normal protein did not contain any cysteic acid. Taken together, our results strongly suggest that sickle protein 4.1 has sustained oxidative damage in vivo. This damage can alter the functional properties of the sickle protein and may be an underlying factor in the myriad of membrane abnormalities reported in sickle erythrocytes.