Analysis of the actin-binding domain of alpha-actinin by mutagenesis and demonstration that dystrophin contains a functionally homologous domain

To define the actin-binding site within the NH2-terminal domain (residues 1-245) of chick smooth muscle alpha-actinin, we expressed a series of alpha-actinin deletion mutants in monkey Cos cells. Mutant alpha-actinins in which residues 2-19, 217-242, and 196-242 were deleted still retained the ability to target to actin filaments and filament ends, suggesting that the actin-binding site is located within residues 20-195. When a truncated alpha-actinin (residues 1-290) was expressed in Cos cells, the protein localized exclusively to filament ends. This activity was retained by a deletion mutant lacking residues 196-242, confirming that these are not essential for actin binding. The actin-binding site in alpha-actinin was further defined by expressing both wild-type and mutant actin-binding domains as fusion proteins in E. coli. Analysis of the ability of such proteins to bind to F-actin in vitro showed that the binding site was located between residues 108 and 189. Using both in vivo and in vitro assays, we have also shown that the sequence KTFT, which is conserved in several members of the alpha-actinin family of actin-binding proteins (residues 36-39 in the chick smooth muscle protein) is not essential for actin binding. Finally, we have established that the NH2-terminal domain of dystrophin is functionally as well as structurally homologous to that in alpha-actinin. Thus, a chimeric protein containing the NH2-terminal region of dystrophin (residues 1-233) fused to alpha-actinin residues 244-888 localized to actin-containing structures when expressed in Cos cells. Furthermore, an E. coli-expressed fusion protein containing dystrophin residues 1-233 was able to bind to F-actin in vitro.     

Structure of the utrophin actin-binding domain bound to F-actin reveals binding by an induced fit mechanism

Utrophin is a large ubiquitously expressed cytoskeletal protein, homologous to dystrophin, the protein disrupted in Duchenne muscular dystrophy. The association of both proteins with the actin cytoskeleton is functionally important and is mediated by a domain at their N termini, conserved in members of the spectrin superfamily, including alpha-actinin, beta-spectrin and fimbrin. We present the structure of the actin-binding domain of utrophin in complex with F-actin, determined by cryo-electron microscopy and helical reconstruction, and a pseudo-atomic model of the complex, generated by docking the crystal structures of the utrophin domain and F-actin into the reconstruction. In contrast to the model of actin binding proposed for fimbrin, the utrophin actin-binding domain appears to associate with actin in an extended conformation. This conformation places residues that are highly conserved in utrophin and other members of the spectrin superfamily at the utrophin interface with actin, confirming the likelihood of this binding orientation. This model emphasises the importance of protein flexibility in modeling interactions and presents the fascinating possibility of a diversity of actin-binding mechanisms among related proteins.     

An atomic model for actin binding by the CH domains and spectrin-repeat modules of utrophin and dystrophin

Utrophin and dystrophin link cytoskeletal F-actin filaments to the plasmalemma. Genetic strategies to replace defective dystrophin with utrophin in individuals with muscular dystrophy requires full characterization of these proteins. Both contain homologous N-terminal actin-binding motifs composed of a pair of calponin-homology (CH) domains (CH1 and CH2) that are connected by spectrin-repeat modules to C-terminal membrane-binding sequences. Here, electron microscopy and 3D reconstruction of F-actin decorated with utrophin and dystrophin actin-binding constructs were performed using Utr261 (utrophin's CH domain pair), Utr416 (utrophin's CH domains and first spectrin-repeat) and Dys246 (dystrophin's CH domain pair). The lozenge-like utrophin CH domain densities localized to the upper surface of actin subdomain 1 and extended azimuthally over subdomain 2 toward subdomains 3 and 4. The cylinder-shaped spectrin-repeat was located at the end of the CH domain pair and was aligned longitudinally along the cleft between inner and outer actin domains, where tropomyosin is present when on thin filaments. The connection between the spectrin-repeat module and the CH domains defined the orientation of CH1 and CH2 on actin. Resolution of utrophin's CH domains and spectrin-repeats permitted docking of crystal structures into respective EM densities, leading to an atomic model where both CH and spectrin-domains bind actin. The CH domain-actin interaction for dystrophin was found to be more complex than for utrophin. Binding assays showed that Utr261 and Utr416 interacted with F-actin as monomers, whereas Dys246 appeared to associate as a dimer, consistent with a bilobed Dys246 structure observed on F-actin in electron microscope reconstructions. One of the lobes was similar in shape, position and orientation to the monomeric CH domains of Utr261, while the other lobe apparently represented a second set of CH domains in the dimeric Dys246. The extensive contact made by dystrophin on actin may be used in vivo to help muscles dissipate mechanical stress from the contractile apparatus to the extracellular matrix.     

Biochemical characterisation of the actin-binding properties of utrophin

Utrophin is a large ubiquitously expressed cytoskeletal protein that is important for maturation of vertebrate neuromuscular junctions. It is highly homologous to dystrophin, the protein defective in Duchenne and Becker muscular dystrophies. Utrophin binds to the actin cytoskeleton via an N-terminal actin-binding domain, which is related to the actin-binding domains of members of the spectrin superfamily of proteins. We have determined the actin-binding properties of this utrophin domain and investigated its binding site on F-actin. An F-actin cosedimentation assay confirmed that the domain binds more tightly to beta-F-actin than to alpha-F-actin and that the full-length utrophin domain binds more tightly to both actin isoforms than a truncated construct, lacking a characteristic utrophin N-terminal extension. Both domain constructs exist in solution as compact monomers and bind to actin as 1:1 complexes. Analysis of the products of partial proteolysis of the domain in the presence of F-actin showed that the N-terminal extension was protected by binding to actin. The actin isoform dependence of utrophin binding could reflect differences at the N-termini of the actin isoforms, thus localising the utrophin-binding site on actin. The involvement of the actin N-terminus in utrophin binding was also supported by competition binding assays using myosin subfragment S1, which also binds F-actin near its N-terminus. Cross-linking studies suggested that utrophin contacts two actin monomers in the actin filament as does myosin S1. These biochemical approaches complement our structural studies and facilitate characterisation of the actin-binding properties of the utrophin actin-binding domain.     

Full-length sequence of the cDNA for human erythroid beta-spectrin

Spectrin is the major molecular consituent of the red cell membrane skeleton. We have isolated overlapping human erythroid beta-spectrin cDNA clones and determined 6773 base pairs of contiguous nucleotide sequence. This includes the entire coding sequence of beta-spectrin. The sequence translates into a 2137 amino acid, 246-kDa peptide. beta-Spectrin is found to consist of three distinct domains. Domain I, at the N terminus, is a 272-amino acid region lacking resemblance to the spectrin repetitive motif. Sequences in this region exhibit striking sequence homology, at both nucleotide and amino acid levels, to the N-terminal "actin-binding" domains of alpha-actinin and dystrophin. Between residues 51 and 270 there is 55% amino acid identity to human dystrophin, with only four single amino acid gaps in alignment. Domain II consists of 17 spectrin repeats. Several sequence variations are observed in typical repeat structure. Homology to alpha-actinin extends beyond domain I into the N-terminal portion of domain II. Domain III, 52 amino acid residues at the C terminus, does not adhere to the spectrin repeat motif. Combining knowledge of spectrin primary structure with previously reported functional studies, it is possible to make several inferences regarding structure/function relationships within the beta-spectrin molecule.     

Specific interaction of the actin-binding domain of dystrophin with intermediate filaments containing keratin 19

Cytokeratins 8 and 19 concentrate at costameres of striated muscle and copurify with the dystrophin-glycoprotein complex, perhaps through the interaction of the cytokeratins with the actin-binding domain of dystrophin. We overexpressed dystrophin's actin-binding domain (Dys-ABD), K8 and K19, as well as closely related proteins, in COS-7 cells to assess the basis and specificity of their interaction. Dys-ABD alone associated with actin microfilaments. Expressed with K8 and K19, which form filaments, Dys-ABD associated preferentially with the cytokeratins. This interaction was specific, as the homologous ABD of betaI-spectrin failed to interact with K8/K19 filaments, and Dys-ABD did not associate with desmin or K8/K18 filaments. Studies in COS-7 cells and in vitro showed that Dys-ABD binds directly and specifically to K19. Expressed in muscle fibers in vivo, K19 accumulated in the myoplasm in structures that contained dystrophin and spectrin and disrupted the organization of the sarcolemma. K8 incorporated into sarcomeres, with no effect on the sarcolemma. Our results show that dystrophin interacts through its ABD with K19 specifically and are consistent with the idea that cytokeratins associate with dystrophin at the sarcolemma of striated muscle.     

Binding sites involved in the interaction of actin with the N-terminal region of dystrophin.

Two actin-binding sites have been identified on human dystrophin by proton NMR spectroscopy of synthetic peptides corresponding to defined regions of the polypeptide sequence. These are Actin-Binding Site 1 (ABS1) located at residues 17-26 and Actin-Binding Site 2 (ABS2) in the region of residues 128-156. Using defined fragments of the actin amino acid sequence, ABS1 has been shown to bind to actin in the region represented by residues 83-117 and ABS2 to the C-terminal region represented by residues 350-375. These dystrophin-binding sites lie on the exposed domain in the actin filament.     

The WW domain of dystrophin requires EF-hands region to interact with beta-dystroglycan.

Skeletal muscle dystrophin is a 427 kDa protein thought to act as a link between the actin cytoskeleton and the extracellular matrix. Perturbations of the dystrophin-associated complex, for example, between dystrophin and the transmembrane glycoprotein beta-dystroglycan, may lead to muscular dystrophy. Previously, the cysteine-rich region and first half of the carboxy-terminal domain of dystrophin were shown to interact with beta-dystroglycan through a stretch of fifteen amino acids at the carboxy-terminus of beta-dystroglycan. This region of dystrophin implicated in binding beta-dystroglycan contains four modular protein domains: a WW domain, two putative Ca2+-binding EF-hand motifs, and a putative zinc finger ZZ domain. The WW domain is a globular domain of 38-40 amino acids with two highly conserved tryptophan residues spaced 20-22 amino acids apart. A subset of WW domains was shown to bind ligands that contain a Pro-Pro-x-Tyr core motif (where x is any amino acid). Here we elucidate the role of the WW domain of dystrophin and surrounding sequence in binding beta-dystroglycan. We show that the WW domain of dystrophin along with the EF-hand motifs binds to the carboxy-terminus of beta-dystroglycan. Through site-specific mutagenesis and in vitro binding assays, we demonstrate that binding of dystrophin to the carboxy-terminus of beta-dystroglycan occurs via a beta-dystroglycan Pro-Pro-x-Tyr core motif. Targeted mutagenesis of conserved WW domain residues reveals that the dystrophin/beta-dystroglycan interaction occurs primarily through the WW domain of dystrophin. Precise mapping of this interaction could aid in therapeutic design.     

Dystrophin and utrophin bind actin through distinct modes of contact

This study was designed to define the molecular epitopes of dystrophin-actin interaction and to directly compare the actin binding properties of dystrophin and utrophin. According to our data, dystrophin and utrophin both bound alongside actin filaments with submicromolar affinities. However, the molecular epitopes involved in actin binding differed between the two proteins. In utrophin, the amino-terminal domain and an adjacent string of the first 10 spectrin-like repeats more fully recapitulated the activities measured for full-length protein. The homologous region of dystrophin bound actin with low affinity and near 1:1 stoichiometry as previously measured for the isolated amino-terminal, tandem (CH) domain. In contrast, a dystrophin construct including a cluster of basic spectrin-like repeats and spanning from the amino terminus through repeat 17, bound actin with properties most similar to full-length dystrophin. Dystrophin and utrophin both stabilized preformed actin filaments from forced depolymerization with similar efficacies but did not appear to compete for binding sites on actin. We also found that dystrophin binding to F-actin was markedly sensitive to increasing ionic strength, although utrophin binding was unaffected. Although dystrophin and utrophin are functionally homologous actin-binding proteins, these results indicate that their respective modes of contact with actin filaments are markedly different. Finally, we reassessed the abundance of dystrophin in striated muscle using full-length protein as the standard and measured greater than 10-fold higher values than previously reported.     

Binding of dystrophin's tandem calponin homology domain to F-actin is modulated by actin's structure

Dystrophin has been shown to be associated in cells with actin bundles. Dys-246, an N-terminal recombinant protein encoding the first 246 residues of dystrophin, includes two calponin-homology (CH) domains, and is similar to a large class of F-actin cross-linking proteins including alpha-actinin, fimbrin, and spectrin. It has been shown that expression or microinjection of amino-terminal fragments of dystrophin or the closely related utrophin resulted in the localization of these protein domains to actin bundles. However, in vitro studies have failed to detect any bundling of actin by either intact dystrophin or Dys-246. We show here that the structure of F-actin can be modulated so that there are two modes of Dys-246 binding, from bundling actin filaments to only binding to single filaments. The changes in F-actin structure that allow Dys-246 to bundle filaments are induced by covalent modification of Cys-374, proteolytic cleavage of F-actin's C-terminus, mutation of yeast actin's N-terminus, and different buffers. The present results suggest that F-actin's structural state can have a large influence on the nature of actin's interaction with other proteins, and these different states need to be considered when conducting in vitro assays.     

A new model for the interaction of dystrophin with F-actin

The F-actin binding and cross-linking properties of skeletal muscle dystrophin-glycoprotein complex were examined using high and low speed cosedimentation assays, microcapillary falling ball viscometry, and electron microscopy. Dystrophin-glycoprotein complex binding to F-actin saturated near 0.042 +/- 0.005 mol/ mol, which corresponds to one dystrophin per 24 actin monomers. Dystrophin-glycoprotein complex bound to F-actin with an average apparent Kd for dystrophin of 0.5 microM. These results demonstrate that native, full-length dystrophin in the glycoprotein complex binds F-actin with some properties similar to those measured for several members of the actin cross-linking super- family of proteins. However, we failed to observe dystrophin- glycoprotein complex-induced cross-linking of F-actin by three different methods, each positively controlled with alpha-actinin. Furthermore, high speed cosedimentation analysis of dystrophin- glycoprotein complex digested with calpain revealed a novel F-actin binding site located near the middle of the dystrophin rod domain. Recombinant dystrophin fragments corresponding to the novel actin binding site and the first 246 amino acids of dystrophin both bound F- actin but with significantly lower affinity and higher capacity than was observed with purified dystrophin-glycoprotein complex. Finally, dystrophin-glycoprotein complex was observed to significantly slow the depolymerization of F-actin, Suggesting that dystrophin may lie along side an actin filament through interaction with multiple actin monomers. These data suggest that although dystrophin is most closely related to the actin cross-linking superfamily based on sequence homology, dystrophin binds F-actin in a manner more analogous to actin side-binding proteins.     

Linking microfilaments to intracellular membranes: the actin-binding and vesicle-associated protein comitin exhibits a mannose-specific lectin activity.

Comitin is a 24 kDa actin-binding protein from Dictyostelium discoideum that is located primarily on Golgi and vesicle membranes. We have probed the molecular basis of comitin's interaction with both actin and membranes using a series of truncation mutants obtained by expressing the appropriate cDNA in Escherichia coli. Comitin dimerizes in solution; its principle actin-binding activity is located between residues 90 and 135. The N-terminal 135 'core' residues of comitin contain a 3-fold sequence repeat that is homologous to several monocotyledon lectins and which retains key residues that determine these lectins' three-dimensional structure and mannose binding. These repeats of comitin appear to mediate its interaction with mannose residues in glycoproteins or glycolipids on the cytoplasmic surface of membrane vesicles from D.discoideum, and comitin can be released from membranes with mannose. Our data indicate that comitin binds to vesicle membranes via mannose residues and, by way of its interaction with actin, links these membranes to the cytoskeleton.     

Occurrence of an actin-inserting domain in tensin.

We have previously described a protein called "insertin" that binds strongly to barbed ends of actin filaments and permits polymerization of actin filaments by insertion of actin monomers between the barbed ends and barbed end-bound insertin. We determined the amino acid sequence of insertin and found that the primary structure of insertin is almost identical to amino acid residues 862 to 1212 of the actin-binding protein tensin.     

Amino acid sequence of a 15 kilodalton actin-binding fragment of turkey gizzard caldesmon: similarity with dystrophin, tropomyosin and the tropomyosin-binding region of troponin T

We have determined the amino acid sequence of a 15 kDa actin-binding fragment of turkey gizzard caldesmon. The 96-residue fragment contains 29 acidic and 29 basic residues, and is predicted to have an extended helical conformation stabilized by numerous internal salt bridges. CaD15 bears some resemblance to dystrophin, tropomyosin and several other proteins, but is most strikingly similar to the tropomyosin-binding segment of troponin T.     

Impacts of dystrophin and utrophin domains on actin structural dynamics: implications for therapeutic design

We have used time-resolved phosphorescence anisotropy (TPA) of actin to evaluate domains of dystrophin and utrophin, with implications for gene therapy in muscular dystrophy. Dystrophin and its homolog utrophin bind to cytoskeletal actin to form mechanical linkages that prevent muscular damage. Because these proteins are too large for most gene therapy vectors, much effort is currently devoted to smaller constructs. We previously used TPA to show that both dystrophin and utrophin have a paradoxical effect on actin rotational dynamics-restricting amplitude while increasing rate, thus increasing resilience, with utrophin more effective than dystrophin. Here, we have evaluated individual domains of these proteins. We found that a "mini-dystrophin," lacking one of the two actin-binding domains, is less effective than dystrophin in regulating actin dynamics, correlating with its moderate effectiveness in rescuing the dystrophic phenotype in mice. In contrast, we found that a "micro-utrophin," with more extensive internal deletions, is as effective as full-length dystrophin in the regulation of actin dynamics. Each of utrophin's actin-binding domains promotes resilience in actin, while dystrophin constructs require the presence of both actin-binding domains and the C-terminal domain for full function. This work supports the use of a utrophin template for gene or protein therapy designs. Resilience of the actin-protein complex, measured by TPA, correlates remarkably well with previous reports of functional rescue by dystrophin and utrophin constructs in mdx mice. We propose the use of TPA as an in vitro method to aid in the design and testing of emerging gene therapy constructs.     

Utrophin binds laterally along actin filaments and can couple costameric actin with sarcolemma when overexpressed in dystrophin-deficient muscle

Dystrophin is widely thought to mechanically link the cortical cytoskeleton with the muscle sarcolemma. Although the dystrophin homolog utrophin can functionally compensate for dystrophin in mice, recent studies question whether utrophin can bind laterally along actin filaments and anchor filaments to the sarcolemma. Herein, we have expressed full-length recombinant utrophin and show that the purified protein is fully soluble with a native molecular weight and molecular dimensions indicative of monomers. We demonstrate that like dystrophin, utrophin can form an extensive lateral association with actin filaments and protect actin filaments from depolymerization in vitro. However, utrophin binds laterally along actin filaments through contribution of acidic spectrin-like repeats rather than the cluster of basic repeats used by dystrophin. We also show that the defective linkage between costameric actin filaments and the sarcolemma in dystrophin-deficient mdx muscle is rescued by overexpression of utrophin. Our results demonstrate that utrophin and dystrophin are functionally interchangeable actin binding proteins, but that the molecular epitopes important for filament binding differ between the two proteins. More generally, our results raise the possibility that spectrin-like repeats may enable some members of the plakin family of cytolinkers to laterally bind and stabilize actin filaments.     

Transgenic mdx mice expressing dystrophin with a deletion in the actin- binding domain display a "mild Becker" phenotype

The functional significance of the actin-binding domain of dystrophin, the protein lacking in patients with Duchenne muscular dystrophy, has remained elusive. Patients with deletions of this domain (domain I) typically express low levels of the truncated protein. Whether the moderate to severe phenotypes associated with such deletions result from loss of an essential function, or from reduced levels of a functional protein, is unclear. To address this question, we have generated transgenic mice that express wild-type levels of a dystrophin deleted for the majority of the actin-binding domain. The transgene derived protein lacks amino acids 45-273, removing 2 of 3 in vitro identified actin interacting sites and part of hinge 1. Examination of the effect of this deletion in mice lacking wild-type dystrophin (mdx) suggests that a functional domain I is not essential for prevention of a dystrophic phenotype. However, in contrast to deletions in the central rod domain and to full-length dystrophin, both of which are functional at only 20% of wild-type levels, proteins with a deletion in domain I must be expressed at high levels to prevent a severe dystrophy. These results are also in contrast to the severe dystrophy resulting from truncation of the COOH-terminal domain that links dystrophin to the extracellular matrix. The mild phenotype observed in mice with domain I-deletions indicates that an intact actin-binding domain is not essential, although it does contribute to an important function of dystrophin. These studies also suggest the link between dystrophin and the subsarcolemmal cytoskeleton involves more than a simple attachment of domain I to actin filaments.     

Comprehensive analysis of all triple helical repeats in beta-spectrins reveals patterns of selective evolutionary conservation

The spectrin superfamily (spectrin, alpha-actinin, utrophin and dystrophin) has in common a triple helical repeating unit of ~106 amino acid residues. In spectrin, alpha and beta chains contain multiple copies of this repeat. beta-spectrin chains contain the majority of binding activities in spectrin and are essential for animal life. Canonical beta-spectrins have 17 repeats; beta-heavy spectrins have 30. Here, the repeats of five human beta-spectrins, plus beta-spectrins from several other vertebrates and invertebrates, have been analysed. Repeats 1, 2, 14 and 17 in canonical beta are highly conserved between invertebrates and vertebrates, and repeat 8 in some isoforms. This is consistent with conservation of critical functions, since repeats 1, 2 and 17 bind alpha-spectrin. Repeats 1 of beta-spectrins are not always detected by SMART or Pfam tools. A profile hidden Markov model of beta-spectrin repeat 1 detects alpha-actinins, but not utrophin or dystrophin. Novel examples of repeat 1 were detected in the spectraplakins MACF1, BPAG1 and plectin close to the actin-binding domain. Ankyrin binds to the C-terminal portion of repeat 14; the high conservation of this entire repeat may point to additional, undiscovered ligand-binding activities. This analysis indicates that the basic triple helical repeat pattern was adapted early in the evolution of the spectrin superfamily to encompass essential binding activities, which characterise individual repeats in proteins extant today.     

The N-Terminal Actin-Binding Tandem Calponin-Homology (CH) Domain of?Dystrophin Is in a Closed Conformation in Solution and When Bound to?F-actin

Deficiency of the vital muscle protein dystrophin triggers Duchenne/Becker muscular dystrophy, but the structure-function relationship of dystrophin is poorly understood. To date, molecular structures of three dystrophin domains have been determined, of which the N-terminal actin-binding domain (N-ABD or ABD1) is of particular interest. This domain is composed of two calponin-homology (CH) domains, which form an important class of ABDs in muscle proteins. A previously determined x-ray structure indicates that the dystrophin N-ABD is a domain-swapped dimer, with each monomer adopting an extended, open conformation in which the two CH domains do not interact. This structure is controversial because it contradicts functional studies and known structures of similar ABDs from other muscle proteins. Here, we investigated the solution conformation of the dystrophin N-ABD using a very simple and elegant technique of pyrene excimer fluorescence. Using the wild-type protein, which contains two cysteines, and the corresponding single-cysteine mutants, we show that the protein is a monomer in solution and is in a closed conformation in which the two CH domains seem to interact, as observed from the excimer fluorescence of pyrene-labeled wild-type protein. Excimer fluorescence was also observed in its actin-bound form, indicating that the dystrophin N-ABD binds to F-actin in a closed conformation. Comparison of the dystrophin N-ABD conformation with other ABDs indicates that the tandem CH domains in general may be monomeric in solution and predominantly occur in closed conformation, whereas their actin-bound conformations may differ.     

Integrators of the cytoskeleton that stabilize microtubules.

Sensory neurodegeneration occurs in mice defective in BPAG1, a gene encoding cytoskeletal linker proteins capable of anchoring neuronal intermediate filaments to actin cytoskeleton. While BPAG1 null mice fail to anchor neurofilaments (NFs), BPAG1/NF null mice still degenerate in the absence of NFs. We report a novel neural splice form that lacks the actin-binding domain and instead binds and stabilizes microtubules. This interaction is functionally important; in mice and in vitro, neurons lacking BPAG1 display short, disorganized, and unstable microtubules defective in axonal transport. Ironically, BPAG1 neural isoforms represent microtubule-associated proteins that when absent lead to devastating consequences. Moreover, BPAG1 can functionally account for the extraordinary stability of axonal microtubules necessary for transport over long distances. Its isoforms interconnect all three cytoskeletal networks, a feature apparently central to neuronal survival.