Mechanisms of cytoskeletal regulation: functional and antigenic diversity in human erythrocyte and brain beta spectrin

A study of human erythrocyte and brain spectrin with particular emphasis on the beta subunits revealed a structural homology but functional dissimilarity between these two molecules. Six monoclonal antibodies raised to human erythrocyte beta spectrin identify three of the four proteolytically defined domains of erythrocyte beta spectrin. Five of these monoclonal antibodies cross-react with human brain spectrin. None of a previously identified set of alpha erythrocyte spectrin monoclonal antibodies [Yurchenco et al: J Biol Chem 257:9102, 1982] reacted with brain spectrin. A domain map generated by limited tryptic digestion shows that brain spectrin is composed of proteolytically resistant domains analogous to erythrocyte spectrin, but the brain protein is more basic. The binding of brain spectrin to erythrocyte ankyrin, both in solution and on erythrocyte IOVs, yielded an association constant approximately 100 time weaker than for erythrocyte spectrin. The binding of azido-calmodulin under native conditions was specific for the erythrocyte beta subunit but was not calcium dependent. In contrast, azido-calmodulin bound only to the alpha subunit of brain spectrin in a calcium-dependent manner. The similarity of structure but modified functional characteristics of the brain and erythrocyte beta spectrins suggest that these proteins serve different cellular roles.     

Beta spectrin bestows protein 4.1 sensitivity on spectrin-actin interactions

The ability of protein 4.1 to stimulate the binding of spectrin to F-actin has been compared by cosedimentation analysis for three avian (erythrocyte, brain, and brush border) and two mammalian (erythrocyte and brain) spectrin isoforms. Human erythroid protein 4.1 stimulated actin binding of all spectrins except the brush border isoform (TW 260/240). These results suggested that the beta subunit determined the protein 4.1 sensitivity of the heterodimer, since all avian alpha subunits are encoded by a single gene. Tissue-specific posttranslational modification of the alpha subunit was excluded by examining the properties of hybrid spectrins composed of the purified alpha subunit from avian erythrocyte or brush border spectrin and the beta subunit of human erythrocyte spectrin. A hybrid composed of avian brush border alpha and human erythroid beta spectrin ran on nondenaturing gels as a discrete band, migrating near human erythroid spectrin tetramers. The actin-binding activity of this hybrid was stimulated by protein 4.1, while either chain alone was devoid of activity. Therefore, although both subunits were required for actin binding, the sensitivity of the spectrin-actin interaction to protein 4.1 is a property uniquely bestowed on the heterodimer by the beta subunit. The singular insensitivity of brush border spectrin to stimulation by erythroid protein 4.1 was also consistent with the absence of proteins in avian intestinal epithelial cells which were immunoreactive with polyclonal antisera sensitive to all of the known avian and human erythroid 4.1 isoforms.     

Identification of a spectrin-like protein in Sertoli cells

Sertoli cells prepared from rats ages 15 and 25 days were shown to contain a spectrin-like protein. Indirect immunofluorescence with monospecific antimouse erythrocyte immunoglobulin G (IgG) and with monospecific antimouse brain spectrin IgG revealed specific staining in Sertoli cells. Both antibodies precipitated two spectrin-like peptides of 240,000 and 235,000 daltons from cells solubilized with octyl glucoside. Proteins from Sertoli cell membranes were separated by electrophoresis on polyacrylamide gels containing sodium dodecyl sulfate and electrophoretically transferred to nitrocellulose membrane. Incubation of nitrocellulose membrane with either of the two antibodies, followed by horseradish peroxidase conjugated to second antibody, revealed only the larger, or alpha, spectrin subunit (Western blots). Both antibodies were used to provide immunoautoradiographic identification of the spectrin-like protein. In this procedure, spectrin and Sertoli cell membranes were shown to compete with [125I]-labeled spectrin from mouse erythrocytes for binding to antimouse erythrocyte spectrin IgG. Finally, two-dimensional proteolytic mapping of the 240,000- and 235,000-dalton peptides demonstrated limited spot homology with rat erythrocyte spectrin. However, subcellular fractions from Sertoli cells all contained a spectrin-like protein showing high homology from fraction to fraction. It is concluded that Sertoli cells contain a spectrin-like protein that is seen in cell fractions prepared by centrifugation, i.e., mitochondria, microsomes, nuclei, cytoplasm, and plasma membranes. Although homology with spectrin from erythrocytes or brain is not seen in peptide maps, the alpha subunit shares antigenic determinants with spectrin from erythrocytes. The beta subunit is believed to be precipitated by antispectrin as the result of binding to the alpha subunit, since the beta subunit shows no detectable antigenic homology with that of spectrin.     

Role of terminal nonhomologous domains in initiation of human red cell spectrin dimerization

Human erythrocyte spectrin is an antiparallel heterodimer comprised of a 280 kDa alpha subunit and a 246 kDa beta subunit which further associates into tetramers in the red cell membrane cytoskeleton. Lateral association of the flexible rodlike monomers involves a multiple-step process that is initiated by a high affinity association near the actin-binding end of the molecule (dimer nucleation site). In this study, recombinant alpha and beta proteins comprising two or four "spectrin type" motifs with and without adjacent, terminal nonhomologous domains were evaluated for their relative contributions to dimer initiation, and the thermodynamic properties of these heterodimer complexes were measured. Sedimentation equilibrium studies showed that in the absence of the heterologous subunit, individual recombinant proteins formed weak homodimers (K(d) > 0.3 mM). When 2-motif (alpha20-21 and beta1-2) and 4-motif (alpha18-21 and beta1-4) recombinants lacking the terminal nonhomologous domains were paired with the complementary protein, high affinity heterodimers were formed in sedimentation equilibrium analysis. Both the alpha20-21/beta1-2 complex and the alpha20-21EF/betaABD1-2 complex showed stoichiometric binding with similar binding affinities (K(d) approximately 10 nM) using isothermal titration calorimetry. The alpha20-21/beta1-2 complex showed an enthalpy of -10 kcal/mol, while the alpha20-21EF/betaABD1-2 complex showed an enthalpy of -13 kcal/mol. Pull-down assays using alpha spectrin GST fusion proteins showed strong associations between all heterodimer complexes in physiological buffer, but all heterodimer complexes were destabilized by the presence of Triton X-100 and other detergents. Complexes lacking the nonhomologous domains were destabilized to a greater extent than complexes that included the nonhomologous domains. The detergent effect appears to be responsible for the apparent essential role of the nonhomologous domains in prior reports. Taken together, our results indicate that the terminal nonhomologous domains do not contribute to dimer initiation nor are they required for formation of high affinity spectrin heterodimers in physiological buffers.     

Phosphatidylserine binding sites in erythroid spectrin: location and implications for membrane stability

The erythrocyte membrane is a composite structure consisting of a lipid bilayer tethered to the spectrin-based membrane skeleton. Two complexes of spectrin with other proteins are known to participate in the attachment. Spectrin has also been shown to interact with phosphatidylserine (PS), a component of the lipid bilayer, which is confined to its inner leaflet. That there may be multiple sites of interaction with PS in the spectrin sequence has been inferred, but they have not hitherto been identified. Here we have explored the interaction of PS-containing liposomes with native alpha- and beta-spectrin chains and with recombinant spectrin fragments encompassing the entire sequences of both chains. We show that both alpha-spectrin and beta-spectrin bind PS and that sites of high affinity are located within 8 of the 38 triple-helical structural repeats which make up the bulk of both chains; these are alpha8, alpha9-10, beta2, beta3, beta4, beta12, beta13, and beta14, and PS affinity was also found in the nonhomologous N-terminal domain of the beta-chain. No other fragments of either chain showed appreciable binding. Binding of spectrin and its constituent chains to mixed liposomes of PS and phosphatidylcholine (PC) depended on the proportion of PS. Binding of spectrin dimers to PS liposomes was inhibited by single repeats containing PS binding sites. It is noteworthy that the PS binding sites in beta-spectrin are grouped in close proximity to the sites of attachment both of ankyrin and of 4.1R, the proteins engaged in attachment of spectrin to the membrane. We conjecture that direct interaction of spectrin with PS in the membrane may modulate its interactions with the proteins and that (considering also the known affinity of 4.1R for PS) the formation of PS-rich lipid domains, which have been observed in the red cell membrane, may be a result.     

Structural and functional studies of interaction between Plasmodium falciparum knob-associated histidine-rich protein (KAHRP) and erythrocyte spectrin

Plasmodium falciparum dramatically modifies the structure and function of the membrane of the parasitized host erythrocyte. Altered membrane properties are the consequence of the interaction of a group of exported malaria proteins with host cell membrane proteins. KAHRP (the knob-associated histidine-rich protein), a member of this group, has been shown to interact with erythrocyte membrane skeletal protein spectrin. However, the molecular basis for this interaction has yet to be defined. In the present study, we defined the binding motifs in both KAHRP and spectrin and identified a functional role for this interaction. We showed that spectrin bound to a 72-amino-acid KAHRP fragment (residues 370-441). Among nine-spectrin fragments, which encompass the entire alpha and beta spectrin molecules (four alpha spectrin and five beta spectrin fragments), KAHRP bound only to one, the alpha N-5 fragment. The KAHRP-binding site within the alpha N-5 fragment was localized uniquely to repeat 4. The interaction of full-length spectrin dimer to KAHRP was inhibited by repeat 4 of alpha spectrin. Importantly, resealing of this repeat peptide into erythrocytes mislocalized KAHRP in the parasitized cells. We concluded that the interaction of KAHRP with spectrin is critical for appropriate membrane localization of KAHRP in parasitized erythrocytes. As the presence of KAHRP at the erythrocyte membrane is necessary for cytoadherence in vivo, our findings have implications for the development of new therapies for mitigating the severity of malaria infection.     

Chick integrin alpha V subunit molecular analysis reveals high conservation of structural domains and association with multiple beta subunits in embryo fibroblasts.

We have cloned and characterized a chick homologue of the human vitronectin receptor alpha subunit (alpha v) whose primary sequence is 83% identical with its human counterpart but less than 40% identical with any other known integrin alpha subunit. Comparison of the chick and human sequences reveals several highly conserved regions, including the cytoplasmic domain. The putative ligand binding domain contains alpha v-specific residues that may contribute to ligand binding specificity. These are concentrated in three regions that are located before and between the first three Ca2+ binding domains. Polyclonal antibodies raised against two peptides deduced from the putative cytoplasmic and extracellular domains of the chick alpha v sequence recognize specifically integrin heterodimers in chick embryo fibroblasts. At least three putative beta subunits coimmunoprecipitate with the chick alpha v subunit. In addition to a protein with the same molecular weight as beta 3 (94K), protein bands of Mr 84K and 110K are also coprecipitated. By successive immunodepletions, we demonstrate that this latter Mr 110K subunit is beta 1, which appears to be one of the alpha v-associated subunits in chick embryo fibroblasts.     

Recombinant extracellular domain of the three major subunits of GABAA receptor show comparable secondary structure and benzodiazepine binding properties

The three most widely expressed subunits of the GABAA receptor are alpha(1), beta(2), and gamma(2) subunits, and the major isoform in the human brain is a pentameric receptor composed of 2alpha(1)2beta(2)1gamma(2). Previously, we overexpressed the extracellular domain Q28-R248 of GABAA receptor alpha(1) subunit. In the present study, the homologous extracellular domains Q25-G243 of GABAA receptor beta(2) subunit and Q40-G273 of gamma(2) subunit were also obtained through overexpression in Escherichia coli. Successful production of recombinant beta(2) and gamma(2) subunit receptor protein domains facilitates the comparison of structural and functional properties of the three subunits. To this end, the secondary structures of the three fragments were measured using CD spectroscopy and the beta-strand contents calculated to be >30%, indicating a beta-rich structure for all three fragments. In addition, the benzodiazepine (BZ)-binding affinity of the recombinant fragments were measured using fluorescence polarization to be 2.16 microM, 3.63 microM, and 1.34 microM for the alpha(1), beta(2), and gamma(2) subunit fragments, respectively, indicating that all three homomeric assemblies, including that of the beta(2) subunit, generally not associated with BZ binding, can bind BZ in the micromolar range. The finding that the BZ binding affinity of these recombinant domains was highest for the gamma(2) subunit and lowest for the beta(2) subunit is consistent with results from previous binding studies using hetero-oligomeric receptors. The present results exemplify the effective approach to characterize and compare the three major subunits of the GABAA receptor, for two of which the overexpression in E. coli is reported for the first time.     

Mapping the domain structure of human erythrocyte adducin

Adducin is a 200-kDa heterodimeric protein associated with the erythrocyte membrane skeleton which binds to Ca2+/calmodulin, promotes binding of spectrin to actin, and is a substrate for protein kinases C and A. Adducin polypeptides can be structurally and functionally divided into two distinct regions. The amino-terminal 39-kDa domain of each subunit is more basic and resistant to proteases than the C-terminal 60-64-kDa domain, which is very sensitive to proteolytic degradation. Two-dimensional peptide map analysis revealed that the 39-kDa protease-resistant domains represent a portion of adducin which is highly conserved between the alpha and beta subunits whereas the protease-sensitive regions are different in each subunit. Comparison of the structural and functional properties of purified 39-kDa domains with intact adducin showed that the 39-kDa domains were not phosphorylated by protein kinases C or A and did not bind to Ca2+/calmodulin or interact with spectrin and actin. This suggests that the protease-sensitive domains may perform the various functions of adducin since these activities were all lacking from the protease-resistant domains. It is also possible that the conserved and variable domains are both required for one or more activities of adducin or that the 39-kDa domains play a role in maintaining the oligomeric state of adducin necessary for interaction of the variable domains with spectrin-actin complexes.     

Contributions of the beta-subunit to spectrin structure and function

The three avian spectrins that have been characterized consist of a common alpha-subunit (240 kD) paired with an isoform-specific beta-subunit from either erythrocyte (220 or 230 kD), brain (235 kD), or intestinal brush border (260 kD). Analysis of avian spectrins, with their naturally occurring "subunit replacement" has proved useful in assessing the relative contribution of each subunit to spectrin function. In this study we have completed a survey of avian spectrin binding properties and present morphometric analysis of the relative flexibility and linearity of various avian and human spectrin isoforms. Evidence is presented that, like its mammalian counterpart, avian brain spectrin binds human erythroid ankyrin with low affinity. Cosedimentation analysis demonstrates that 1) avian erythroid protein 4.1 stimulates spectrin-actin binding of both mammalian and avian erythrocyte and brain spectrins, but not the TW 260/240 isoform, 2) calpactin I does not potentiate actin binding of either TW 260/240 or brain spectrin, and 3) erythrocyte adducin does not stimulate the interaction of TW 260/240 with actin. In addition, a morphometric analysis of rotary-shadow images of spectrin isoforms, individual subunits, and reconstituted complexes from isolated subunits was performed. This analysis revealed that the overall flexibility and linearity of a given spectrin heterodimer and tetramer is largely determined by the intrinsic rigidity and linearity of its beta-spectrin subunit. No additional rigidity appears to be imparted by noncovalent associations between the subunits. The scaled flexural rigidity of the most rigid spectrin analyzed (human brain) is similar to that reported for F-actin.     

Topological analysis of the pyridine nucleotide transhydrogenase of Escherichia coli using proteolytic enzymes

The pyridine nucleotide transhydrogenase of Escherichia coli has an alpha 2 beta 2 structure (alpha: Mr, 54,000; beta: Mr, 48,700). Hydropathy analysis of the amino acid sequences suggested that the 10 kDa C-terminal portion of the alpha subunit and the N-terminal 20-25 kDa region of the beta subunit are composed of transmembranous alpha-helices. The topology of these subunits in the membrane was investigated using proteolytic enzymes. Trypsin digestion of everted cytoplasmic membrane vesicles released a 43 kDa polypeptide from the alpha subunit. The beta subunit was not susceptible to trypsin digestion. However, it was digested by proteinase K in everted vesicles. Both alpha and beta subunits were not attacked by trypsin and proteinase K in right-side out membrane vesicles. The beta subunit in the solubilized enzyme was only susceptible to digestion by trypsin if the substrates NADP(H) were present. NAD(H) did not affect digestion of the beta subunit. Digestion of the beta subunit of the membrane-bound enzyme by trypsin was not induced by NADP(H) unless the membranes had been previously stripped of extrinsic proteins by detergent. It is concluded that binding of NADP(H) induces a conformational change in the transhydrogenase. The location of the trypsin cleavage sites in the sequences of the alpha and beta subunits were determined by N- and C-terminal sequencing. A model is proposed in which the N-terminal 43 kDa region of the alpha subunit and the C-terminal 30 kDa region of the beta subunit are exposed on the cytoplasmic side of the inner membrane of E. coli. Binding sites for pyridine nucleotide coenzymes in these regions were suggested by affinity chromatography on NAD-agarose columns.     

Interaction between the subunits of human erythrocyte spectrin using a fluorescence probe

Fluorescence labeling of spectrin subunits was performed with N-(1-anilinonaphthyl-4)maleimide (ANM) to study the interaction between alpha and beta subunits. The fluorescence anisotropy of both ANM alpha and ANM beta increased linearly with the addition of nonfluorescent beta or alpha subunit, and saturated at a protein ratio about 1, indicating that 1 mol alpha subunit binds to 1 mol beta subunit with high affinity in vitro. Furthermore, this binding seemed to be reversible, because the anisotropy value decreased when an excess fo nonfluorescent alpha was added to the ANM alpha/beta mixture. The anisotropy of ANM alpha attained a maximum level within l min after addition of the same quantity of nonfluorescent beta at 12 degrees C, and the anisotropy of this mixture decreased rapidly when an excess of nonfluorescent alpha was added. These findings suggested that both the binding process of beta to ANM alpha and the dissociation step of ANM alpha from the ANM alpha-beta complex were quite rapid. The results obtained here imply that dynamic interaction between alpha and beta subunits of spectrin should be taken into account in understanding the role of the spectrin molecule in the cytoskeletal mesh.     

A method for assessing the reassembly of a multisubunit glycoprotein by western blotting

The Western blot procedure has been adapted to detect the reassembly of a two-subunit glycoprotein, urinary human chorionic gonadotropin (hCG), directly on the nitrocellulose. This glycoprotein is composed of two nonidentical subunits, alpha and beta. A simple procedure using immunoblotting has been developed to detect reassembly of the monomers to dimer. Three monoclonal antibodies were required for the development of this method: A109, which binds the alpha subunit or hCG; B105, which binds the beta subunit or hCG; and B107, specific for the intact hCG dimer. The alpha subunit and beta subunit of hCG were each electrophoresed and transferred to nitrocellulose, and the transfer was then incubated with the appropriate complementary subunit; reassembly of the dimer was determined by the binding of the monoclonal antibody B107. Evidence that the reassembly occurs directly on the nitrocellulose comes from the fact that B107 immunoreactivity is detected at the molecular weight position of the subunit and not at the dimer molecular weight. A genetically expressed recombinant form of the alpha subunit was also tested for its ability to recombine with the opposite subunit to produce the dimer. The recombinant alpha subunit was determined to have additional carbohydrate which interfered with the binding of the beta subunit. N-Glycanase digestion of the recombinant alpha subunit produced a form which, when incubated with the beta subunit, did recombine on the nitrocellulose and could be recognized by B107.     

Cytoplasmic and transmembrane domains of integrin beta 1 and beta 3 subunits are functionally interchangeable

Integrin beta subunits combine with specific sets of alpha subunits to form functional adhesion receptors. The structure and binding properties of integrins suggest the presence of domains controlling at least three major functions: subunit association, ligand binding, and cytoskeletal interactions. To more carefully define structure/function relationships, a cDNA construct consisting of the extracellular domain of the avian beta 1 subunit and the cytoplasmic and transmembrane domains of the human beta 3 subunit was prepared and expressed in murine 3T3 cells. The resulting chimeric beta 1/3 subunit formed heterodimers with alpha subunits from the beta 1 subfamily, could not interact with alpha IIb from the beta 3 subfamily, was targeted to focal contacts, and formed functional complexes within the focal contacts. A second cDNA construct was prepared that coded for an avian beta 1 subunit without a transmembrane or cytoplasmic domain. This subunit was not found in association with an accompanying alpha subunit, nor was it found expressed on the cell surface. Instead, it accumulated in vesicles within the cytoplasm and was eventually shed from the cell. The results from studies of the behavior of these two cDNA constructs demonstrate that the transmembrane and cytoplasmic domains play no role in alpha subunit selection, that the cytoplasmic domain of beta 3 is capable of functioning in the context of alpha subunits with which it is not normally paired, and that both integrin subunits must be membrane associated for normal assembly and transport to cell surface adhesive structures.     

Mechanisms of cytoskeletal regulation: modulation of membrane affinity in avian brush border and erythrocyte spectrins

The spectrins isolated from chicken erythrocytes and chicken intestinal brush border, TW260/240, share a common alpha subunit and a tissue-specific beta subunit. The ability of these related proteins to bind human erythrocyte inside out vesicles (IOVs) and human erythrocyte ankyrin in vitro have been quantitatively compared with human erythrocyte spectrin. Chicken erythrocyte spectrin binds human IOVs and human ankyrin with affinities nearly identical to that for human erythrocyte spectrin. TW260/240 does not significantly bind to either IOVs or ankyrin. These results demonstrate a remarkable tissue preservation of ankyrin-binding capacity, even between diverse species, and confirm the role of the avian beta-spectrins in modulating this functionality. Avian brush border spectrin may represent a unique spectrin which serves primarily as a filament cross-linker and which does not interact strongly with membrane-associated proteins.     

Three distinct forms of type 2A protein phosphatase in human erythrocyte cytosol

Two type 2A protein phosphatases, phosphatases I (Mr = 180,000) and III (Mr = 177,000), were purified to near homogeneity from human erythrocyte cytosol. Phosphatase I was composed of alpha (34 kDa), beta (63 kDa), and delta (74 kDa) subunits in a ratio of 1:1:1. Phosphatase III comprised alpha, beta, and gamma (53 kDa) subunits in the same ratio. Heparin-Sepharose column chromatography converted most of phosphatase I and 20% of phosphatase III into alpha 1 beta 1 which were indistinguishable from phosphatase IV (Usui, H., Kinohara, N., Yoshikawa, K., Imazu, M., Imaoka, T., and Takeda, M. (1983) J. Biol. Chem. 258, 10455-10463). The catalytic subunit alpha and the beta subunit of phosphatases I, III, and IV displayed identical V8 and papain peptide maps, respectively, while the peptide maps of the alpha, beta, gamma, and delta subunits were clearly distinct. The molar ratio of phosphatases I, III, and IV in erythrocyte cytosol was estimated to be 6:1:14. Comparison of molecular activities of alpha, alpha 1 beta 1, alpha 1 beta 1 delta 1, and alpha 1 beta 1 gamma 1 revealed that beta suppressed phosphorylase and P-H2B histone phosphatase activities of alpha but stimulated the P-H1 histone phosphatase activity, and delta suppressed all the phosphatase activities of alpha 1 beta 1. The gamma subunit stimulated the P-histone phosphatase activity of alpha 1 beta 1 but inhibited the phosphorylase and P-spectrin phosphatase activities. The beta subunit increased the Mg2+ or Mn2+ requirement for P-H2B histone phosphatase activity of alpha, an effect which was counteracted by delta. The effects of heparin, H1 histone, protamine, and polylysine on the phosphorylase phosphatase activity of phosphatases I, III, IV, and alpha were described and discussed in connection with the functions of the subunits.     

In vitro proteolysis of brain spectrin by calpain I inhibits association of spectrin with ankyrin-independent membrane binding site(s).

This report demonstrates that specific proteolysis of brain spectrin by a calcium-dependent protease, calpain I, abolishes association of brain spectrin with the ankyrin-independent binding site(s) in brain membranes. Calpain I cleaves the beta subunit of spectrin at the N-terminal end leaving a 218-kDa fragment and cleaves the alpha subunit in the midregion to produce 150- and 130-kDa fragments. Calpain-proteolyzed spectrin almost completely loses the capacity to displace binding of intact spectrin to membranes. Spectrin digested by calpain I under conditions that almost completely destroyed membrane-binding remained associated as a tetramer and retained about 60% of the ability to associate with actin filaments. Cleavage of spectrin occurred at sites distinct from the membrane-binding site which is located on the beta subunit since the isolated 218-kDa fragment of the beta subunit as well as a reconstituted complex of alpha and 218-kDa beta subunit fragment partially regained binding activity. Moreover, cleavage of the alpha subunit alone reduced the affinity of spectrin for membranes by 2-fold. A consequence of distinct sites for calpain I cleavage and membrane-binding is that calpain I can digest spectrin while spectrin is complexed with other proteins and therefore has the potential to mediate disassembly of a spectrin-actin network from membranes.     

Brain ankyrin. Purification of a 72,000 Mr spectrin-binding domain

Polypeptides of Mr = 190,000-220,000 that cross-react with erythrocyte ankyrin were detected in immunoblots of membranes from pig lens, pig brain, and rat liver. The cross-reacting polypeptides from brain were cleaved by chymotrypsin to fragments of Mr = 95,000 and 72,000 which are the same size as fragments obtained with erythrocyte ankyrin. The brain 72,000 Mr fragment associated with erythrocyte spectrin, and the binding occurred at the same site as that of erythrocyte ankyrin 72,000 Mr fragment since (a) brain 72,000 Mr fragment was adsorbed to erythrocyte spectrin-agarose and (b) 125I-labeled erythrocyte spectrin bound to brain 72,000 Mr fragment following transfer of the fragment from a sodium dodecyl sulfate gel to nitrocellulose paper, and this binding was displaced by erythrocyte ankyrin 72,000 Mr fragment. Brain 72,000 Mr fragment was purified about 400-fold by selective extraction and by continuous chromatography on columns attached in series containing DEAE-cellulose followed by erythrocyte spectrin coupled to agarose, and finally hydroxylapatite. The brain 72,000 Mr fragment was not derived from contaminating erythrocytes since peptide maps of pig brain and pig erythrocyte 72,000 Mr fragments were distinct. The amount of brain 72,000 Mr fragment was estimated as 0.28% of membrane protein or 39 pmol/mg based on radioimmunoassay with 125I-labeled brain fragment and antibody against erythrocyte ankyrin. Brain spectrin tetramer was present in about the same number of copies (30 pmol/mg of membrane protein) based on densitometry of Coomassie blue-stained sodium dodecyl sulfate gels. The binding site on brain spectrin for both brain and erythrocyte ankyrin 72,000 Mr fragments was localized by electron microscopy to the midregion of spectrin tetramers about 90 nM from the near end and 110 nM from the far end. These studies demonstrate the presence in brain membranes of a protein closely related to erythrocyte ankyrin, and are consistent with a function of the brain ankyrin as a membrane attachment site for brain spectrin.     

A beta-spectrin isoform from Drosophila (beta H) is similar in size to vertebrate dystrophin

Spectrins are a major component of the membrane skeleton in many cell types where they are thought to contribute to cell form and membrane organization. Diversity among spectrin isoforms, especially their beta subunits, is associated with diversity in cell shape and membrane architecture. Here we describe a spectrin isoform from Drosophila that consists of a conventional alpha spectrin subunit complexed with a novel high molecular weight beta subunit (430 kD) that we term beta H. The native alpha beta H molecule binds actin filaments with high affinity and has a typical spectrin morphology except that it is longer than most other spectrin isoforms and includes two knoblike structures that are attributed to a unique domain of the beta H subunit. Beta H is encoded by a different gene than the previously described Drosophila beta-spectrin subunit but shows sequence similarity to beta-spectrin as well as vertebrate dystrophin, a component of the membrane skeleton in muscle. By size and sequence similarity, dystrophin is more similar to this newly described beta-spectrin isoform (beta H) than to other members of the spectrin gene family such as alpha-spectrin and alpha- actinin.     

Ankyrin and synapsin: spectrin-binding proteins associated with brain membranes

Brain membranes contain an actin-binding protein closely related in structure and function to erythrocyte spectrin. The proteins that attach brain spectrin to membranes are not established, but, by analogy with the erythrocyte membrane, may include ankyrin and protein 4.1. In support of this idea, proteins closely related to ankyrin and 4.1 have been purified from brain and have been demonstrated to associate with brain spectrin. Brain ankyrin binds with high affinity to the spectrin beta subunit at the midregion of spectrin tetramers. Brain ankyrin also has binding sites for the cytoplasmic domain of the erythrocyte anion channel (band 3), as well as for tubulin. Ankyrins from brain and erythrocytes have a similar domain structure with protease-resistant domains of Mr = 72,000 that contain spectrin-binding activity, and domains of Mr = 95,000 (brain ankyrin) or 90,000 (erythrocyte ankyrin) that contain binding sites for both tubulin and the anion channel. Brain ankyrin is present at about 100 pmol/mg membrane protein, or about twice the number of copies of spectrum beta chains. Brain ankyrin thus is present in sufficient amounts to attach spectrin to membranes, and it has the potential to attach microtubules to membranes as well as to interconnect microtubules with spectrin-associated actin filaments. Another spectrin-binding protein has been purified from brain membranes, and this protein cross-reacts with erythrocyte 4.1. Brain 4.1 is identical to the membrane protein synapsin, which is one of the brain's major substrates for cAMP-dependent and Ca/calmodulin-dependent protein kinases with equivalent physical properties, immunological cross-reaction, and peptide maps. Synapsin (4.1) is present at about 60 pmol/mg membrane protein, and thus is a logical candidate to regulate certain protein linkages involving spectrin.