Preparation of red-cell-membrane cytoskeletal constituents and characterisation of protein 4.1

Diseases of the Nucleoskeleton

The nucleus is separated from the cytosol by the nuclear envelope, which is a double lipid bilayer composed of the outer nuclear membrane and the inner nuclear membrane. The intermediate filament proteins lamin A, lamin B, and lamin C form a network underlying the inner nuclear membrane. This proteinaceous network provides the nucleus with its strength, rigidity, and elasticity. Positioned within the inner nuclear membrane are more than 150 inner nuclear membrane proteins, many of which interact directly with lamins and require lamins for their inner nuclear membrane localization. Inner nuclear membrane proteins and the nuclear lamins define the nuclear lamina. These inner nuclear membrane proteins have tissue-specific expression and diverse functions including regulating cytoskeletal organization, nuclear architecture, cell cycle dynamics, and genomic organization. Loss or mutations in lamins and inner nuclear membrane proteins cause a wide spectrum of diseases. Here, I will review the functions of the well-studied nuclear lamina proteins and the diseases associated with loss or mutations in these proteins. © 2016 American Physiological Society. Compr Physiol 6:1655-1674, 2016.

TRPing on the lung endothelium: calcium channels that regulate barrier function

Rises in cytosolic calcium are sufficient to initiate the retraction of endothelial cell borders and to increase macromolecular permeability. Although endothelial cell biologists have recognized the importance of shifts in cytosolic calcium for several decades, only recently have we gained a rudimentary understanding of the membrane calcium channels that change cell shape. Members of the transient receptor potential family (TRP) are chief among the molecular candidates for permeability-coupled calcium channels. Activation of calcium entry through store-operated calcium entry channels, most notably TRPC1 and TRPC4, increases lung endothelial cell permeability, as does activation of calcium entry through the TRPV4 channel. However, TRPC1 and TRPC4 channels appear to influence the lung extraalveolar endothelial barrier most prominently, whereas TRPV4 channels appear to influence the lung capillary endothelial barrier most prominently. Thus, phenotypic heterogeneity in ion channel expression and function exists within the lung endothelium, along the arterial-capillary-venous axis, and is coupled to discrete control of endothelial barrier function.

Two distinct domains of protein 4.1 critical for assembly of functional nuclei in vitro

Protein 4.1R, a multifunctional structural protein, acts as an adaptor in mature red cell membrane skeletons linking spectrin-actin complexes to plasma membrane-associated proteins. In nucleated cells protein 4.1 is not associated exclusively with plasma membrane but is also detected at several important subcellular locations crucial for cell division. To identify 4.1 domains having critical functions in nuclear assembly, 4.1 domain peptides were added to Xenopus egg extract nuclear reconstitution reactions. Morphologically disorganized, replication deficient nuclei assembled when spectrin-actin-binding domain or NuMA-binding C-terminal domain peptides were present. However, control variant spectrin-actin-binding domain peptides incapable of binding actin or mutant C-terminal domain peptides with reduced NuMA binding had no deleterious effects on nuclear reconstitution. To test whether 4.1 is required for proper nuclear assembly, 4.1 isoforms were depleted with spectrin-actin binding or C-terminal domain-specific antibodies. Nuclei assembled in the depleted extracts were deranged. However, nuclear assembly could be rescued by the addition of recombinant 4.1R. Our data establish that protein 4.1 is essential for nuclear assembly and identify two distinct 4.1 domains, initially characterized in cytoskeletal interactions, that have crucial and versatile functions in nuclear assembly.

Host cell and malarial targets for docetaxel (Taxotere) during the erythrocytic development of Plasmodium falciparum

The microtubular stabilizing agent docetaxel (Taxotere) is known to inhibit the intraerythrocytic development of Plasmodium falciparum. To investigate the mechanism(s) of inhibition, we analyzed the structural organization of the mitotic spindle by immunofluorescence and electron microscopy. When 30 microM docetaxel was applied for five hours on ring forms, alterations in the mitotic spindles leading to abnormal nuclear divisions were observed. At the trophozoite- and schizont-stage, docetaxel pulses prevent mitosis by stabilizing microtubular structures associated with the mitotic apparatus, giving abnormal spindles. However, this inhibition did not interfere with parasite DNA synthesis indicating the absence of a checkpoint that couples exit from mitosis with proper spindle assembly as observed in higher eukaryotic cells. In parallel, intraerythrocytic concentration of docetaxel was measured in parasitized erythrocytes, after incubation of cells with 3H-docetaxel for five hours. It was found to be 14-fold increased at the ring-stage of infected erythrocytes compared to normal ones, 170-fold increased at the trophozoite-stage and 1,500-fold increased at the schizont-stage. Our data show that, even though the overall intracellular concentration of docetaxel is low in docetaxel-pulsed rings, the agent might be sufficient to disturb the spindle organization. However, the existence of targets for docetaxel other than mitotic spindle microtubules, i.e. erythrocyte membrane components could interfere with mitotic spindle formation.

Detection of dimers of dimers of human leukocyte antigen (HLA)-DR on the surface of living cells by single-particle fluorescence imaging

The technique of single-particle fluorescence imaging was used to investigate the oligomeric state of MHC class II molecules on the surface of living cells. Cells transfected with human leukocyte antigen (HLA)-DR A and B genes were labeled at saturation with a univalent probe consisting of Fab coupled to R-phycoerythrin. Analysis of the intensities of fluorescent spots on the cell surface revealed the presence of single and double particles consistent with the simultaneous presence of HLA-DR heterodimers and dimers of dimers. The proportion of double particles was lower at 37 degrees C than at 22 degrees C, suggesting that the heterodimers and dimers of dimers exist in a temperature-dependent equilibrium. These results are discussed in the context of a possible role for HLA-DR dimers of dimers in T cell receptor-MHC interactions. The technique is validated by demonstrating that fluorescence imaging can distinguish between dimers and tetramers of human erythrocyte spectrin deposited from solution onto a solid substrate. The methodology will have broad applicability to investigation of the oligomeric state of immunological and other membrane-bound receptors in living cells.

Cytoskeletal protein binding kinetics at planar phospholipid membranes

It has been hypothesized that nonspecific reversible binding of cytoskeletal proteins to lipids in cells may guide their binding to integral membrane anchor proteins. In a model system, we measured desorption rates k(off) (off-rates) of the erythrocyte cytoskeletal proteins spectrin and protein 4.1 labeled with carboxyfluorescein (CF), at two different compositions of planar phospholipid membranes (supported on glass), using the total internal reflection/fluorescence recovery after photobleaching (TIR/FRAP) technique. The lipid membranes consisted of either pure phosphatidylcholine (PC) or a 3:1 mixture of PC with phosphatidylserine (PS). In general, the off-rates were not single exponentials and were fit to a combination of fast, slow, and irreversible fractions, reported both separately and as a weighted average. By a variation of TIR/FRAP, we also measured equilibrium affinities (the ratio of surface-bound to bulk protein concentration) and thereby calculated on-rates, k(on). The average off-rate of CF-4.1 from PC/PS (approximately 0.008/s) is much slower than that from pure PC (approximately 1.7/s). Despite the consequent increase in equilibrium affinity at PC/PS, the on-rate at PC/PS is also substantially decreased (by a factor of 40) relative to that at pure PC. The simultaneous presence of (unlabeled) spectrin tends to substantially decrease the on-rate (and the affinity) of CF-4.1 at both membrane types. Similar experiments for CF-spectrin alone showed much less sensitivity to membrane type and generally faster off-rates than those exhibited by CF-4.1. However, when mixed with (unlabeled) 4.1, both the on-rate and off-rate of CF-spectrin decreased drastically at PC/PS (but not PC), leading to a somewhat increased affinity. Clearly, changes in affinity often involve countervailing changes in both on-rates and off-rates. In many of these studies, the effect of varying ionic strength and bulk concentrations was examined; it appears that the binding is an electrostatic attraction and is far from saturation at the concentrations employed. These results and the techniques implemented carry general implications for understanding the functional role of nonspecific protein binding to cellular lipid membranes.

Cytoskeleton-membrane connections in the human erythrocyte membrane: band 4.1 binds to tetrameric band 3 protein

Band 4.1 provides, besides ankyrin, the main linkage between the erythrocyte membrane and its cytoskeleton. Its predominant binding sites in the membrane are located on the glycophorins. However, the cytoplasmic domain of band 3 can also bind band 4.1. We have studied which of the different band 3 oligomers observed (monomers, dimers, tetramers) can act as band 4.1 binding sites, by equilibrium sedimentation experiments on mixtures of purified band 3 and dye-labelled band 4.1 in solutions of a nonionic detergent. At low molar ratios of band 4.1 and band 3, the sedimentation equilibrium distributions obtained could all be perfectly fitted assuming that only two dye-labelled particles were present: uncomplexed band 4.1 and a complex formed between one band 4.1 molecule and one band 3 tetramer. The presence of small amounts of complexes containing band 3 monomers or dimers could not be completely ruled out but is unlikely. On the other hand, stabilized band 3 dimers effectively bound band 4.1. At higher molar band 4.1/band 3 ratio, the band 3 tetramer apparently could bind up to at least four band 4.1 molecules. The band 4.1/band 3 tetramer complex was found to be unstable. The results described, together with those reported previously, point at a prominent role of tetrameric band 3 in ligand binding.

Modular organization of the PDZ domains in the human discs-large protein suggests a mechanism for coupling PDZ domain-binding proteins to ATP and the membrane cytoskeleton

The human homologue (hDIg) of the Drosophila discs-large tumor suppressor (DIg) is a multidomain protein consisting of a carboxyl-terminal guanylate kinase-like domain, an SH3 domain, and three slightly divergent copies of the PDZ (DHR/GLGF) domain. Here have examined the structural organization of the three PDZ domains of hDIg using a combination of protease digestion and in vitro binding measurements. Our results show that the PDZ domains are organized into two conformationally stable modules one (PDZ, consisting of PDZ domains 1 and 2, and the other (PDZ) corresponding to the third PDZ domain. Using amino acid sequencing and mass spectrometry, we determined the boundaries of the PDZ domains after digestion with endoproteinase Asp-N, trypsin, and alpha-chymotrypsin. The purified PDZ1+2, but not the PDZ3 domain, contains a high affinity binding site for the cytoplasmic domain of Shaker-type K+ channels. Similarly, we demonstrate that the PDZ1+2 domain can also specifically bind to ATP. Furthermore, we provide evidence for an in vivo interaction between hDIg and protein 4.1 and show that the hDIg protein contains a single high affinity protein 4.1-binding site that is not located within the PDZ domains. The results suggest a mechanism by which PDZ domain-binding proteins may be coupled to ATP and the membrane cytoskeleton via hDlg.

Neutral polymers elicit, and antibodies to spectrin, band 4.1 protein and cytoplasmic domain of band 3 protein inhibit the concanavalin A-mediated agglutination of human erythrocytes

Concanavalin A (Con A) is known to agglutinate human erythrocytes if the cells are pre-treated with a proteinase or neuraminidase. We report that untreated cells can also be made to agglutinate with the lectin if the lectin-bound cells are treated with anti-Con A antibodies, or if a neutral polymer such as serum albumin, polyvinylpyrrolidone or Ficoll is added. Thus, Con A falls in the category of 'incomplete' lectins. The polymer induces Con A-agglutinability without altering the receptor number, or deformability of the cells. If the polymer is sequestered within erythrocyte ghosts, Con A is unable to agglutinate them; but the presence of the polymer only on the outer surface (as in intact cells) or on both the surfaces permits agglutinability. Thus, the site of the polymer effect resides on the outer surface of the membrane. The polymer, however, is unable to induce agglutinability in erythrocyte vesicles, whose membrane lacks skeletal proteins. The result suggests a positive role for the membrane skeleton in the process of agglutination brought about by the polymer, as is true also for the agglutination of proteinase-treated cells. In order to obtain detailed information on the proteins participating in agglutination, monospecific antibodies to spectrins, band 4.1 protein, ankyrin and the cytoplasmic domain of band 3 protein were internalized in erythrocytes. It is found that anti-spectrin and anti-band 3 cytoplasmic domain, but not their Fab's, inhibit the Con A-mediated agglutinability partially, and anti-4.1 antibodies, as well as the Fab's, inhibit the agglutinability substantially. Anti-ankyrin, however, was without any effect. The results confirm a positive role for the membrane skeleton in the Con A-mediated agglutination of normal erythrocytes in the presence of a neutral polymer, or in proteinase treated cells. We also provide evidence for requirement of Mg-ATP in the agglutination process.

Identification of the protein 4.1 binding interface on glycophorin C and p55, a homologue of the Drosophila discs-large tumor suppressor protein

Protein 4.1 is the prototype of a family of proteins that include ezrin, talin, brain tumor suppressor merlin, and tyrosine phosphatases. All members of the protein 4.1 superfamily share a highly conserved N-terminal 30-kDa domain whose biological function is poorly understood. It is believed that the attachment of the cytoskeleton to the membrane may be mediated via this 30-kDa domain, a function that requires formation of multiprotein complexes at the plasma membrane. In this investigation, synthetically tagged peptides and bacterially expressed proteins were used to map the protein 4.1 binding site on human erythroid glycophorin C, a transmembrane glycoprotein, and on human erythroid p55, a palmitoylated peripheral membrane phosphoprotein. The results show that the 30-kDa domain of protein 4.1 binds to a 12-amino acid segment within the cytoplasmic domain of glycophorin C and to a positively charged, 39-amino acid motif in p55. Sequences similar to this charged motif are conserved in other members of the p55 superfamily, including the Drosophila discs-large tumor suppressor protein. Our data provide new insights into how protein 4.1, glycophorin C, p55, and their non-erythroid homologues, interact with the cytoskeleton to exert their physiological effects.

Cloning and characterization of hdlg: the human homologue of the Drosophila discs large tumor suppressor binds to protein 4.1

The Drosophila discs large tumor suppressor protein, dlg, has been shown to regulate the growth of imaginal discs during embryogenesis [Woods, D. F. & Bryant, P. J. (1991) Cell 66, 451-464]. We cloned and sequenced the complete cDNA for a human B-lymphocyte 100-kDa protein that shares 60% amino acid identity with dlg. This human homologue of Drosophila discs large (hdlg) contains a C-terminal domain homologous to the known guanylate kinases, a src homology 3 region motif, and three dlg homology repeats. Two nonhomologous domains that can contain in-frame insertions result in at least four alternatively spliced isoforms of hdlg. Several hdlg RNA transcripts are widely distributed in human and murine tissues, and the protein is localized to regions of cell-cell contact. Protein 4.1, the defining member of a family that includes talin and merlin/schwannomin, has the same cellular localization as hdlg, and two sites within hdlg associate in vitro with the 30-kDa N-terminal domain of protein 4.1.

Reversible binding kinetics of a cytoskeletal protein at the erythrocyte submembrane

Reversible binding among components of the cellular submembrane cytoskeleton and reversible binding of some of these components with the plasma membrane likely play a role in nonelastic morphological changes and mechanoplastic properties of cells. However, relatively few studies have been devoted to investigating directly the kinetic aspects of the interactions of individual components of the membrane skeleton with the membrane. The experiments described here investigated whether one component of the erythrocyte membrane cytoskeleton, protein 4.1, binds to its sites on the membrane reversibly and if so, whether the different 4.1-binding sites display distinct kinetic behavior. Protein 4.1 is known to stabilize the membrane and to mediate the attachment of spectrin filaments to the membrane. Protein 4.1 previously has been shown to bind to integral membrane proteins band 3, glycophorin C, and to negatively charged phospholipids. To examine the kinetic rates of dissociation of carboxymethyl fluorescein-labeled 4.1 (CF-4.1) to the cytofacial surface of erythrocyte membrane, a special preparation of hemolyzed erythrocyte ghosts was used, in which the ghosts became flattened on a glass surface and exposed their cytofacial surfaces to the solution through a membrane rip in a distinctive characteristic pattern. This preparation was examined by the microscopy technique of total internal reflection/fluorescence recovery after photobleaching (TIR/FRAP). Four different treatments were employed to help identify which membrane binding sites gave rise to the multiplicity of observed kinetic rates. The first treatment, the control, stripped off the native spectrin, actin, 4.1, and ankyrin. About 60% of the CF-4.1 bound to this control binded irreversibly (dissociation time > 20 min), but the remaining approximately 40% binded reversibly with a range of residency times averaging approximately 3 s. The second treatment subjected these stripped membranes to trypsin, which presumably removed most of the band 3. CF-4.1 binded significantly less to these trypsinized membranes and most of the decrease was a loss of the irreversibly binding sites. The third treatment simply preserved the native 4.1 and ankyrin. CF-4.1 binded less to this sample too, and the loss involved both the irreversible and reversible sites. The fourth treatment blocked the gycophorin C sites on the native 4.1-stripped membranes with an antibody. CF-4.1 again binded less to this sample than to a nonimmune serum control, and almost all of the decrease is a loss of irreversible sites. These rest suggest that 1) protein 4.1 binds to membrane or submembrane sites at least in part reversibly ; 2) the most reversible sites are probably not proteinaceous and not glycophorin C, but possibly are phospholipids (especially phosphatidylserine); and 3) TIWRFRAP can successfully examine the fast reversible dynamics of cytoskeletal components binding to biological membranes.

Fluorescence studies of spectrin and its subunits

To better understand the solution structure of spectrin, the environment of its tryptophan residues have been examined by fluorescence spectroscopy. The spectra and the extent of quenching by several quenching agents have been determined for intact spectrin and its alpha and beta subunits. The arsenal of quenchers used in the study represented both hydrophilic and hydrophobic species including anionic, cationic and neutral compounds. Effects on spectrin fluorescence of ethanol and ionic strength, which extend and/or rigidify spectrin, and of glycerol, which is commonly used in electron microscopy of the protein, have also been assessed in the presence and absence of quenchers. Most of the tryptophans of spectrin are either internally quenched or are sequestered, hindering the approach of hydrophilic quenching agents. Both the spectral shape and the extent of quenching by acrylamide indicate that some tryptophans of the beta subunit are slightly more exposed in the isolated chain than in the dimer. Similar effects on spectra and on quenching of the intact dimer and of the isolated beta chain are seen when the ionic strength is reduced. Ethanol and glycerol reduce spectrin tryptophan accessibility to 2-p-toluidinyl napthalene-6-sulfonic acid (TNS). It therefore appears that low ionic strength, alpha-beta association and neutral solute (or lowered dielectric constant) all induce a similar, but modest conformational change in the domain structure. The extent of TNS binding is not increased by lowering the ionic strength, suggesting that the expansion and/or stiffening of the molecule in low electrolyte solution does not involve exposure of significant numbers of hydrophobic sites.

The role of inositol phospholipids in the association of band 4.1 with the human erythrocyte membrane

Band 4.1 is a major protein of the erythrocyte membrane skeleton. It promotes the binding of spectrin to F-actin and may anchor the skeletal network to the plasma membrane via its association with integral membrane proteins. Here, we have investigated the involvement of inositol phospholipids in the binding of band 4.1 to erythrocyte membranes using membrane vesicles stripped of all peripheral proteins at alkaline pH. Trypsinization of these vesicles allows the discrimination of two classes of band 4.1 binding sites: trypsin-sensitive sites (60-65% of the total), largely or exclusively on band 3, and trypsin-resistant sites (35-40% of the total), composed, at least in part, of the glycophorins. ATP depletion or activation of erythrocyte phosphoinositol phospholipase C led to a reduction in membrane phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P2] content by 20-70% in different experiments. The resulting decrease of band 4.1 binding to vesicles by was variable, but averaged about 15-20%. The same treatments led to an average decrease in the band 4.1 binding capacity of trypsinized vesicles of 55%. Since this is equivalent to a 20% decrease in the binding capacity of non-trypsinized vesicles (consistent with the above result), it indicates that PtdIns(4,5)P2 regulates the binding of band 4.1 only to trypsin-resistant binding sites (and to only a subset of these) accounting for about 15-20% of total band 4.1 binding sites on membranes. We found that hydrolysis of > 95% of PtdIns(4,5)P2 with exogenous phospholipase C-delta (PLC delta) resulted in no further decrease in band 4.1 binding to vesicles than did hydrolysis of 65-70% of PtdIns(4,5)P2 which is accessible to erythrocyte phosphoinositol phospholipase C. This suggests that only 65-70% of total membrane PtdIns(4,5)P2 is involved in regulating band 4.1 binding. Significantly, the pool of PtdIns(4,5)P2 involved is the same pool which can be hydrolysed by erythrocyte phosphoinositol phospholipase C, and which has been shown to be metabolically labile in erythrocytes. The membrane binding capacity for band 4.1 found in this study (averaging 1000 micrograms/mg vesicle protein) is considerably higher than that found in previous studies. The results are consistent with the existence of a binding site for band 4.1 on each copy of the major transmembrane proteins (band 3 and the glycophorins). These results provide new insights into the involvement of membrane inositol phospholipids in cytoskeletal-membrane interactions.

Purification of erythrocyte protein 4.1 by selective interaction with inositol hexaphosphate

Protein 4.1 is a multifunctional structural protein occupying a strategic position in the erythrocyte membrane. It is present in the erythrocyte membrane skeleton and in many nonerythroid cells. This report describes a novel method for purifying this protein based on its selective interaction with inositol hexaphosphate dimagnesium tetrapotassium salt. This interaction was discovered in the course of chromatography of high-salt extract of inside-out membrane vesicles on Procion orange MX-2R-Sepharose. The new procedure is simple and selective and produces protein 4.1 with better yield than that obtained with a previously published procedure. The purified protein 4.1 has the same immunoreactivity and the same alpha-chymotryptic digest profile as protein 4.1 purified by published methods and is fully functional in enhancing the interaction between F-actin and spectrin dimers.

Characterization of calcium binding to spectrins

Calcium binding to brain and erythrocyte spectrins was studied at physiological ionic strength by a calcium overlay assay and aqueous two-phase partitioning. When the spectrins were immobilized on nylon membranes by slot blotting, the overlay assay showed that even though both spectrins bound 45Ca2+, the brain protein displayed much greater affinity for calcium ions than erythrocyte spectrin did. Since the observed binding was weaker than that displayed by calmodulin under similar conditions, the overlay assay results indicated that the binding must be weaker than 1 microM. The phase partition experiments showed that there are at least two sites for calcium on brain spectrin and that calcium binding to one of these sites is reduced significantly by magnesium ions. From the partition isotherm, the dissociation constants were estimated as 50 microM for the Mg(2+)-independent site and 150 microM for the Mg(2+)-dependent site. The phase partition results also showed that erythrocyte spectrin bound calcium ions at least 1 order of magnitude weaker. By examining calcium binding to slot-blotted synthetic peptides, we identified two binding sites in brain spectrin. One mapped to the second putative calcium binding site (EF-hand) in alpha-spectrin and the other to the 36 amino acid residue long insert in domain 11. In addition, a tryptic fragment derived from the C-terminal of erythrocyte alpha-spectrin, which contained the two postulated EF-hands, also bound calcium. These findings suggest that the calcium signal system may also involve direct binding of calcium to spectrin beside known calcium modulators such as calmodulin and calpain.(ABSTRACT TRUNCATED AT 250 WORDS)

Fluorescence quenching of spectrin and other red cell membrane cytoskeletal proteins. Relation to hydrophobic binding sites

The intrinsic fluorescence of spectrin is strongly quenched by low concentrations of 2-bromostearate. This results from binding at a series of hydrophobic sites. Analysis of dynamic fluorescence quenching by acrylamide, iodide and caesium ions, separately and in conjunction with 2-bromostearate, leads to the conclusion that most of the tryptophan side-chains are exposed to solvent. The sites at which the fatty-acid-quenched tryptophans are located apparently interact with the lipid bilayer in the cell, as judged by quenching by bromostearate dissolved in the lipid phase. A minor proportion of the side-chains in native spectrin give rise to sharp proton magnetic resonance signals, indicative of segmental mobility; these chain elements contain some tryptophan residues, as revealed by weak downfield signals from the heterocyclic ring protons. These signals are not appreciably perturbed by stearic acid or by phosphatidylserine liposomes, suggesting that the hydrophobic binding sites are not in mobile chain elements. By contrast with a series of globular proteins which, with the exception of serum albumins, show little or no quenching by 2-bromostearate, the peripheral red cell membrane skeletal proteins ankyrin (and its spectrin-binding domain), protein 4.1 and (to a lesser extent) actin show evidence of a high affinity for the hydrophobic ligand and may, like spectrin, interact directly with the bilayer in situ.

Modulation of protein 4.1 binding to inside-out membrane vesicles by phosphorylation

The effect of phosphorylation on the binding of protein 4.1 to erythrocyte inside-out vesicles was investigated. Protein 4.1 was phosphorylated with casein kinase A, protein kinase C, and cAMP-dependent protein kinase. An analysis of the phosphopeptides generated by alpha-chymotryptic and tryptic digestion indicates these kinases phosphorylate similar as well as distinct domains within protein 4.1. All three enzymes catalyze the phosphorylation to varying degrees of the 46-, 16-, and 8-10-kDa fragments derived from limited chymotryptic cleavage. In addition, casein kinase A phosphorylates a 24-kDa domain, whereas protein kinase C phosphorylates a 30-kDa domain. Protein 4.1 phosphorylated by casein kinase A and protein kinase C, but not cAMP-dependent protein kinase, exhibits a reduced binding to KI-extracted inside-out vesicles. On the other hand, phosphorylation of inside-out vesicles by casein kinase A does not affect their ability to bind protein 4.1. The inside-out vesicles, however, inhibit the phosphorylation of protein 4.1 by casein kinase A and protein kinase C, but not by cAMP-dependent protein kinase. These results suggest that casein kinase A and protein kinase C may modulate the binding of protein 4.1 to the membrane by phosphorylation of specific domains of the cytoskeletal protein. Since the 30-kDa domain has been suggested as a membrane-binding site, that phosphorylation by protein kinase C reduces the binding of protein 4.1 to inside-out vesicles is perhaps not surprising. On the other hand, the role of the casein kinase A substrate 24-kDa domain in membrane binding has not been established and needs to be examined.(ABSTRACT TRUNCATED AT 250 WORDS)

Purification and properties of p103, a novel 103-kDa component of postsynaptic densities

A 103-kDa protein present in membrane cytoskeletal preparations from bovine brain has been identified. We have purified this protein to greater than 95% homogeneity using gel filtration and ion-exchange chromatography. This protein, p103, is an asymmetric dimer in dilute solution and has two major variants that can be distinguished by isoelectric focussing, pI 5.60 and 5.75. Using subcellular fractionation, it is most enriched in postsynaptic densities. Immunolocalization with anti-p103-specific antibodies reveals that it is confined to the dendrites and perikarya; it is apparently absent from spinal cord axons. It coextracts from brain membrane-skeletal preparations with brain spectrin and actin, but in vitro, it does not interact with them.

Characteristics of spectrin-induced leakage of extruded, phosphatidylserine vesicles

At neutral pH spectrin induces modest leakage of trapped calcein from reverse-phase or extruded, but not sonicated, vesicles composed of phosphatidylserine, but not phosphatidylcholine. The extent of leakage from extruded vesicles is not or is only slightly affected by magnesium ions at a physiological concentration or calcium ions at a greater than physiological concentration, respectively. In addition to accounting for several previously discrepant observations on the lytic effects of spectrin, these findings indicate that some proteins like spectrin may destabilize vesicles with low curvature more readily than vesicles of high curvature, in contrast to certain amphiphilic peptides. 60% less leakage is induced from phosphatidylserine vesicles by heat-denatured than by native spectrin. In contrast, both trypsin- and subtilisin-treated spectrins, if sufficiently digested, induce several-fold more leakage than undigested spectrin. Since spectrin prepared either by 1 M Tris dissociation of Triton-extracted cytoskeletons or by low ionic strength extraction of ghosts released the same amounts of calcein from vesicles of various compositions, these effects are unlikely to reflect artifacts of spectrin preparation. Furthermore, spectrin is unlikely to promote leakage in vivo, since vesicles composed of phosphatidylserine, cholesterol and/or phosphatidylethanolamine, which constitute the lipid composition of the inner monolayer of the red cell membrane, did not leak on addition of spectrin, whereas vesicles composed of phosphatidylserine and phosphatidylcholine, did leak in the presence of spectrin.

Long-term intercalation of residual hemin in erythrocyte membranes distorts the cell

The effect of long-term incubation of residual globin-free hemin on whole red blood cell and isolated cytoskeletal proteins was studied. Hemin at concentrations found in pathological red cells was inserted to fresh erythrocytes. Increased hemolysis developed in the hemin-containing cells after a few days at 37 degrees C and after about four weeks at 4 degrees C. Since lipid and hemoglobin peroxidation did not depend on the presence of hemin, time-dependent effects on the cytoskeleton proteins were studied. Observations were: (1) spectrin and protein 4.1 exhibited a time-dependent increasing tendency to undergo hemin-induced peroxidative crosslinking. (2) The ability of the serum proteins, albumin and hemopexin, to draw hemin from spectrin, actin and protein 4.1 decreased with time of incubation with hemin. These results were attributed to time-dependent hemin-induced denaturation of the cytoskeletal proteins. Albumin taken as a control for physiological hemin trap was unaffected by hemin. Small amounts of hemo-spectrin (2-5%) were analyzed in circulating normal cells, and this in vivo hemo-spectrin also failed to release hemin. It was concluded that slow accumulation of hemin, a phenomenon increased in pathological cells, is a toxic event causing erythrocyte destruction.

Heat-induced alterations in monkey erythrocyte membrane phospholipid organization and skeletal protein structure and interactions

Rhesus monkey erythrocytes were subjected to heating at 50 degrees C for 5-15 min, and the heat-induced effects on the membrane structure were ascertained by analysing the membrane phospholipid organization and membrane skeleton dynamics and interactions in the heated cells. Membrane skeleton dynamics and interactions were determined by measuring the Tris-induced dissociation of the Triton-insoluble membrane skeleton (Triton shells), the spectrin-actin extractability at low ionic strength, spectrin self-association and spectrin binding to normal monkey erythrocyte membrane inside-out vesicles (IOVs). The Tris-induced Triton shell dissociation and spectrin-actin extractability were markedly decreased by the erythrocyte heating. Also, the binding of the heated erythrocyte membrane spectrin-actin with the IOVs was much smaller than that observed with the normal erythrocyte spectrin-actin. Further, the spectrin structure was extensively modified in the heated cells, as compared to the normal erythrocytes. Transbilayer phospholipid organization was ascertained by employing bee venom and pancreatic phospholipases A2, fluorescamine, and Merocyanine 540 as the external membrane probes. The amounts of aminophospholipids hydrolysed by phospholipases A2 or labeled by fluorescamine in intact erythrocytes considerably increased after subjecting them to heating at 50 degrees C for 15 min. Also, the fluorescent dye Merocyanine 540 readily stained the 15-min-heated cells but not the fresh erythrocytes. Unlike these findings, the extent of aminophospholipid hydrolysis in 5-min-heated cells by phospholipases A2 depended on the incubation time. While no change in the membrane phospholipid organization could be detected in 10 min, prolonged incubations led to the increased aminophospholipid hydrolysis. Similarly, fluorescamine failed to detect any change in the transbilayer phospholipid distribution soon after the 5 min heating, but it labeled greater amounts of aminophospholipids in the 5-min-heated cells, as compared to normal cells, after incubating them for 4 h at 37 degrees C. These results have been discussed to analyse the role of membrane skeleton in maintaining the erythrocyte membrane phospholipid asymmetry. It has been concluded that both the ATP-dependent aminophospholipid pump and membrane bilayer-skeleton interactions are required to maintain the transbilayer phospholipid asymmetry in native erythrocyte membrane.

Hemin-promoted peroxidation of red cell cytoskeletal proteins

Hemin-induced crosslinking of the erythrocyte membrane proteins was analyzed at three levels: (i) whole membranes, (ii) integrated or dissociated cytoskeletons, and (iii) isolated forms of the three main cytoskeletal proteins, spectrin, actin, and protein 4.1. Addition of H2O2 and hemoglobin to resealed membranes from without did not affect any of the membrane proteins. Hemin that can transport across the membrane induced, in the presence of H2O2, crosslinking of protein 4.1 and spectrin. Both free hemin and hemoglobin added with H2O2 induced crosslinking of integer cytoskeletons and mixtures of isolated cytoskeletal proteins, but hemin was always more active. Of the three major cytoskeletal proteins, spectrin and protein 4.1 were most active while the participation of actin was only minor. The yield of crosslinked products was increased in all reaction mixtures with pH, with an apparent pK above 9.0. Replacement of H2O2 by phenylhydrazine and tert-butyl hydroperoxide resulted in crosslinking of the same proteins, but with lower activity than H2O2. Bityrosines, which were identified by their specific fluorescence emission characteristics, were formed in reaction mixtures containing hemin and hydrogen peroxide and either spectrin or protein 4.1, but not actin. On the basis of fact that bityrosines were revealed only in reaction mixtures that produced protein adducts, formation of intermolecular bityrosines was analyzed to be involved in crosslinking of the cytoskeletal proteins. Since the levels of membrane-intercalated hemin are correlated with aggregation of membrane proteins, it is suggested that the peroxidative properties of hemin are responsible for its toxicity.

Membrane skeleton-bilayer interaction is not the major determinant of membrane phospholipid asymmetry in human erythrocytes

Transbilayer phospholipid distribution, membrane skeleton dissociation/association, and spectrin structure have been analysed in human erythrocytes after subjecting them to heating at 50 degrees C for 15 min. The membrane skeleton dissociation/association was determined by measuring the Tris-induced dissociation of Triton-insoluble membrane skeletons (Triton shells), the spectrin-actin extractability under low ionic conditions, and the binding of spectrin-actin with normal erythrocyte membrane inside-out vesicles (IOVs). The spectrin structure was ascertained by measuring the spectrin dimer-to-tetramer ratio as well as the spectrin tryptophan fluorescence. Both the Tris-induced Triton shell dissociation and the spectrin-actin extractability under low ionic conditions were considerably reduced by the heat treatment. Also, the binding of heated erythrocyte spectrin-actin to IOVs was significantly smaller than that observed with the normal cell spectrin-actin. Further, the quantity of spectrin dimers was appreciably increased in heat-treated erythrocytes as compared to the normal cells. This change in the spectrin dimer-to-tetramer ratio was accompanied by marked changes in the spectrin tryptophan fluorescence. In spite of these heat-induced alterations in structure and bilayer interactions of the membrane skeleton, the inside-outside glycerophospholipid distribution remained virtually unaffected in the heat-treated cells, as judged by employing bee venom and pancreatic phospholipase A2, fluorescamine and Merocyanine 540 as the external membrane probes. These results strongly indicate that membrane bilayer-skeleton interaction is not the major factor in determining the transbilayer phospholipid asymmetry in human erythrocyte membrane.

Composition of the ternary protein complex of the red cell membrane cytoskeleton

The red cell membrane skeletal network is constructed from actin, spectrin and protein 4.1 in a molar ratio of actin subunits/spectrin heterodimer/protein 4.1 of 2:1:1. This represents saturation of the actin filaments, since incubation with extraneous spectrin and protein 4.1 leads to no binding of additional spectrin, either to the inner surface of ghost membranes or to lipid-free membrane cytoskeletons. Partial extraction of spectrin from the membrane is accompanied by release of actin under all conditions. Regardless of the proportion of spectrin extracted, the molar ratio of spectrin dimers/actin subunits is constant at 1:2. This is not the result of release or cooperative breakdown of whole lattice junctions from the network, for the number of actin filaments, judged by capacity to nucleate polymerisation of added G-actin, remains unchanged even when as much as 60% of the total spectrin has been lost. A similar 1:2:1 stoichiometry characterises the complex formed when G-actin is allowed to polymerise in the presence of varying amounts of spectrin and protein 4.1. When this complex is treated with the depolymerising agent, 1 M guanidine hydrochloride, it breaks down into smaller units of the same stoichiometry. After cross-linking these can be recovered from a gel-filtration column. Complexes prepared starting from G-actin appear to be much more stable than those formed when spectrin and protein 4.1 are bound to F-actin.

The spectrin-actin junction of erythrocyte membrane skeletons

High-resolution electron microscopy of erythrocyte membrane skeletons has provided striking images of a regular lattice-like organization with five or six spectrin molecules attached to short actin filaments to form a sheet of five- and six-sided polygons. Visualization of the membrane skeletons has focused attention on the (spectrin)5,6-actin oligomers, which form the vertices of the polygons, as basic structural units of the lattice. Membrane skeletons and isolated junctional complexes contain four proteins that are stable components of this structure in the following ratios: 1 mol of spectrin dimer, 2-3 mol of actin, 1 mol of protein 4.1 and 0.1-0.5 mol of protein 4.9 (numbers refer to mobility on SDS gels). Additional proteins have been identified that are candidates to interact with the junction, based on in vitro assays, although they have not yet been localized to this structure and include: tropomyosin, tropomyosin-binding protein and adducin. The spectrin-actin complex with its associated proteins has a key structural role in mediating cross-linking of spectrin into the network of the membrane skeleton, and is a potential site for regulation of membrane properties. The purpose of this article is to review properties of known and potential constituent proteins of the spectrin-actin junction, regulation of their interactions, the role of junction proteins in erythrocyte membrane dysfunction, and to consider aspects of assembly of the junctions.

Membrane protein lesions in erythrocytes with Heinz bodies

We studied Heinz body-containing erythrocytes with three different unstable hemoglobins: Nottingham, Brockton, and unclassified. We demonstrated two classes of membrane protein defects in unstable hemoglobin-containing cells (UH-RBCs), a defect of the spectrin-depleted inside-out vesicle (UH-IOV), and a defect of spectrin (UH-spectrin) itself. The composition of UH-IOVs is the same as control with respect to quantity of ankyrin and proportion inside-out. However, UH-IOVs bind even less spectrin than IOVs derived from sickle erythrocytes (SS-IOVs), suggesting a severe functional defect in the ankyrin of UH-RBCs (UH-ankyrin). Further evidence that UH-ankyrin is abnormal is demonstrated by the virtual absence of ankyrin in isotonic membrane shells of UH-RBCs (UH-shells), and abnormal mobility and decreased binding of the 72-kD (spectrin-binding) alpha-chymotryptic fragment of UH-ankyrin (UH-72-kD) to control spectrin. All UH-RBC membranes were spectrin-deficient (60% of control). In addition, spectrin isolated from UH-RBCs (UH-spectrin) was abnormal in two respects: (a) presence of a fast-moving band on nondenaturing polyacrylamide gels of both 0 degree C and 37 degrees C extracts, and (b) decreased binding to actin in the presence of protein 4.1. UH-spectrin did exhibit normal self-association, binding to IOVs and binding to actin in the absence of protein 4.1. This pattern of normal and abnormal spectrin functions has been described for spectrin subjected to mild diamide oxidation, suggesting the role of oxidation is the pathogenesis of membrane defect(s) of erythrocytes with abnormal hemoglobins.

Identification and partial purification of ABGP205, an integral membrane glycoprotein from brain that binds ankyrin

1. A procedure was devised that allows the membrane-skeletal proteins brain spectrin and ankyrin to be extracted selectively from a membrane-skeletal preparation, together with some actin, an Mr-103,000 protein and a population of glycoproteins. 2. Ankyrin-binding activities of the glycoproteins were investigated by affinity chromatography. We detected only one, Mr 205,000, that binds ankyrin and is prevented from binding by the cytoplasmic domain of Band 3, the established erythrocyte-membrane-binding site for ankyrin. The Mr-205,000 glycoprotein, designated ABGP205, may be a candidate for a membrane-binding site for ankyrin.

Immunohistochemical localization of a membrane-associated, 4.1-like protein in the rat visual cortex during postnatal development

Expression and localization of a membrane-associated protein, an analog of erythrocyte protein 4.1, in the visual cortex were immunohistochemically studied in the rat, ranging in age from newborn to adult. In the adult, dendrites and somas of layer V pyramidal cells were stained by the antiprotein 4.1 antibody. In most of these immunoreactive neurons, the plasma membrane seemed to be preferentially stained. Neurons located in layers II and III of the cortex were only faintly stained, and those in layers IV and VI were not stained. At birth, the immunoreactivity was already present in pyramidal cells located in the upper part of the cortical subplate. Immature neurons located in the cortical plate were not stained by the antibody, suggesting that the 4.1-like protein is expressed only in the neurons that have differentiated or are differentiating. At postnatal days 2-8, immunoreactive neurons were dramatically increased in layers V and VI and intense labeling was seen at the apical dendrites of layer V pyramidal cells. Most of the stained processes of these and other neurons showed a sign of rapid dendritic growth, i.e., growth cones and filopidia. At days 10-17, the basal dendrites of pyramidal cells in layers II and III became detectable, although still slender. At days 20-37, these dendrites in layers II, III, and V became intensely immunoreactive, and dendritic spines were visualized by the antibody. Throughout all the ages, axons of neurons and neuroglia were not stained by the antibody. Also, most of the neurons in layer IV of the cortex were not immunoreactive. These results suggest that the 4.1-like protein is abundantly expressed in growing parts of the dendrites and spines. A hypothesis that this protein may play a role in synaptic plasticity in the developing visual cortex is discussed.

Spectrin and related molecules

This review begins with a complete discussion of the erythrocyte spectrin membrane skeleton. Particular attention is given to our current knowledge of the structure of the RBC spectrin molecule, its synthesis, assembly, and turnover, and its interactions with spectrin-binding proteins (ankyrin, protein 4.1, and actin). We then give a historical account of the discovery of nonerythroid spectrin. Since the chicken intestinal form of spectrin (TW260/240) and the brain form of spectrin (fodrin) are the best characterized of the nonerythroid spectrins, we compare these molecules to RBC spectrin. Studies establishing the existence of two brain spectrin isoforms are discussed, including a description of the location of these spectrin isoforms at the light- and electron-microscope level of resolution; a comparison of their structure and interactions with spectrin-binding proteins (ankyrin, actin, synapsin I, amelin, and calmodulin); a description of their expression during brain development; and hypotheses concerning their potential roles in axonal transport and synaptic transmission.

Identification of a new 84/82 kDa calmodulin-binding protein, which also interacts with actin filaments, tubulin and spectrin, as synapsin I

A new 84/82 kDa calmodulin-binding protein, which also interacts with actin filaments, tubulin and spectrin, was purified from the bovine synaptosomal membrane. The binding of calmodulin to this protein was Ca2+-dependent, and was inhibited by trifluoperazine, the association constant being calculated to be 2.2 X 10(6) M-1. Maximally, 1 mol of calmodulin bound to 1 mol of the purified protein. This protein was phosphorylated by both kinase II (Ca2+- and calmodulin-dependent kinase) and cyclic AMP-dependent kinase. In addition, antibody against this protein was demonstrated to have an immunological crossreactivity with synapsin I in the synaptosomal membrane.

Calspectin (fodrin or nonerythroid spectrin)-actin interaction: a possible involvement of 4.1-related protein

The calspectin/actin complex extracted from the bovine brain membrane crosslinks F-actin, resulting in the increasing viscosity of F-actin determined by low-shear viscometry. We demonstrated the presence of a protein factor in this complex, which regulated the calspectin-F-actin interaction in a Ca2+- and calmodulin-dependent manner. Erythrocyte protein 4.1, but not synapsin I, mimics the function of this brain factor using a reconstitution system including purified calspectin, calmodulin and F-actin. In the brain complex, the Mr 120,000 and the Mr 80,000/77,000 polypeptides were detected to crossreact with anti-protein 4.1 antibody.

The 240-kDa subunit of human erythrocyte spectrin binds calmodulin at micromolar calcium concentrations

The binding of the isolated alpha-subunit of human erythrocyte spectrin to calmodulin is demonstrated by partitioning in aqueous two-phase systems. The affinity of the alpha-subunit for calmodulin is slightly higher than that of the spectrin dimer, whereas the beta-subunit interacts only very weakly. The binding is in all cases calcium-dependent and is abolished on addition of chlorpromazine. At an ionic strength close to physiological conditions, about 1 microM free calcium is required to induce maximum binding of calmodulin to spectrin dimer.

Purification of erythrocyte band 4.1 and other cytoskeletal components using hydroxyapatite-Ultrogel

An improved method for purifying erythrocyte band 4.1, the protein which mediates the interaction between spectrin and actin, has been developed. The new procedure, using adsorption chromatography on hydroxylapatite crystals immobilized within a crosslinked agarose gel (HA-Ultrogel), is simple and reproducibly provides a high yield of band 4.1 which is essentially free of protein kinase. Other components eluted from the hydroxylapatite matrix include band 4.9, ankyrin, and a 35,000-Da polypeptide that appears to be glyceraldehyde-3-phosphate dehydrogenase that remains bound to the erythrocyte membrane in 150 mM NaCl.

Effect of spectrin dimer on actin polymerization

Spectrin dimer is shown to influence the polymerization behaviour of actin. The polymerization of both Mg2+- and Ca2+-actin is regulated by an enhancement in the rate of nucleation and a fragmentation of preformed actin filaments. In addition, spectrin decreases the critical concentration of Ca2+-actin but not that of Mg2+-actin. This suggests that the two types of actin may differ in their interaction with spectrin dimer probably due to the different conformations. Band 4.1 elevates the effects of spectrin under non-equilibrium conditions but its contribution is less at steady state.