Atomic structure of the actin:DNase I complex

Myosin filament ATPase is enhanced by intramolecularly cross-linked actin

Reaction of rabbit skeletal muscle F-actin with the lysine-directed photolabile cross-linker, N-5-azido-2-nitrobenzoyloxy succinimide was limited to Lysine-328 and Lysine-326, with Lysine-328 being labelled to a greater extent. Photolysis of the modified actin enhanced the actin-activated MgATPase activity of filamentous scallop myosin 3-4-fold more than unmodified actin, without affecting calcium sensitivity. Unphotolysed modified actin behaved as untreated actin, indicating that photolysis was essential for the effect. The actin-activated ATPase of filamentous rabbit myosin was similarly increased by photolysed N-5-azido-2-nitrobenzoyloxy succinimide-modified actin. After photolysis in either the monomeric (G-) or filamentous (F-) form, N-5-azido-2-nitrobenzoyloxy succinimide-modified actin moved as a monomeric (42 kDa) species on SDS gels, and depolymerized and polymerized readily, demonstrating that any cross-linking event produced by photolysis must be intramolecular. In contrast to the substantial increase in actin-activated ATPase activity observed when photolysed ANB-NOS-modified actin was added to filamentous myosin, the enhancement was not observed with the soluble HMM and S-1 fragments of myosin. Photolysed modified actin showed only poor movement on a rabbit HMM-coated surface in vitro motility assays. These results can be explained if the internally cross-linked G-actin subunits which comprise only a fraction of the actin population, either weaken the actin-actin contacts or have an increased affinity for myosin.

Electron diffraction of helical particles

The development of low-dose electron cryo-microscopy has provided the means to see structural details to better than 10 A resolution in helical structures. The application of techniques of image analysis to micrographs can yield accurate phases, but not amplitudes with which to generate three-dimensional maps of the structure. Electron diffraction can provide reliable amplitudes, which can be combined with the phases from the images. In order to collect amplitude data, two problems have to be overcome: the pattern should be obtained from a large well ordered sample of particles, and the inelastic background should be properly subtracted. In this paper, we present three simple methods to produce rafts of helical particles. Using these methods we have obtained electron diffraction patterns from TMV (with data out to 0.28 nm), TMV protein stacked disks (with data out to 0.3 nm) and bacterial flagellar filaments (with data out to 0.5 nm). In addition, we describe the algorithms used to extract the amplitudes from the diffraction patterns.

Cytoplasmic dynein and actin-related protein Arp1 are required for normal nuclear distribution in filamentous fungi

Cytoplasmic dynein is a multisubunit, microtubule-dependent mechanochemical enzyme that has been proposed to function in a variety of intracellular movements, including minus-end-directed transport of organelles. Dynein-mediated vesicle transport is stimulated in vitro by addition of the Glued/dynactin complex raising the possibility that these two complexes interact in vivo. We report here that a class of phenotypically identical mutants of the filamentous fungus Neurospora crassa are defective in genes encoding subunits of either cytoplasmic dynein or the Glued/dynactin complex. These mutants, defined as ropy, have curled hyphae with abnormal nuclear distribution. ro-1 encodes the heavy chain of cytoplasmic dynein, while ro-4 encodes an actin-related protein that is a probable homologue of the actin-related protein Arpl (formerly referred to as actin-RPV or centractin), the major component of the glued/dynactin complex. The phenotypes of ro-1 and ro-4 mutants suggest that cytoplasmic dynein, as well as the Glued/dynactin complex, are required to maintain uniform nuclear distribution in fungal hyphae. We propose that cytoplasmic dynein maintains nuclear distribution through sliding of antiparallel microtubules emanating from neighboring spindle pole bodies.

Mobile segments in rabbit skeletal muscle F-actin detected by 1H nuclear magnetic resonance spectroscopy

Polymerization of actin by increasing the ionic strength leads to a quenching of almost all 1H NMR signals. Surprisingly, distinct signals with relatively small line widths can still be observed in actin filaments (F-actin) indicating the existence of mobile, NMR visible residues in the macromolecular structure. The intensity of the F-actin spectrum is much reduced if one replaces Mg2+ with Ca2+, and a moderate reduction of the signal intensity can also be obtained by increasing the ionic strength. These results can be explained in a two-state model of the actin promoters with a M- (mobile) state and a I- (immobile) state in equilibrium. In the M-state a number of residues in the actin promoter are mobile and give rise to observable NMR signals. This equilibrium is shifted towards the I-state specifically by replacing Mg2+ with Ca(2+)-ions and unspecifically by addition of monovalent ions such as K+. The binding of phalloidin to its high-affinity site in the filaments does not influence the equilibrium between M- and I-state. Phalloidin itself is completely immobilized in F-actin, its exchange with the solvent being slow on the NMR time scale.

DNase-I-like enzyme from the carp liver--inhibition by muscle and endogenous actin

1. DNase-I-like activity occurs in the carp (Cyprinus carpio) liver cytosol (supernatant 105,000 g). 2. The enzyme resembles DNase I from bovine pancreas in respect to the molecular mass (approximately 31 kDa), pH (7.4) and ion requirements (Mg2+, Ca2+) and the ability to degrade native as well as denatured DNA. 3. As judged by comparison of DNase zymograms obtained after native- and SDS-PAGE, the enzyme occurs in the three molecular forms of similar molecular weight and different charges. 4. All these forms are inhibited by rabbit skeletal muscle actin as well as by endogenous actin isolated from the carp liver cytosol. 5. DNase from the carp liver cytosol does not interact with the antibodies directed against DNase I from bovine pancreas and against DNase I from the rat and bovine parotid glands.

Clostridial ADP-ribosylating toxins: effects on ATP and GTP-binding proteins

The actin cytoskeleton appears to be as the cellular target of various clostridial ADP-ribosyltransferases which have been described during recent years. Clostridium botulinum C2 toxin, Clostridium perfringens iota toxin and Clostridium spiroforme toxin ADP-ribosylate actin monomers and inhibit actin polymerization. Clostridium botulinum exoenzyme C3 and Clostridium limosum exoenzyme ADP-ribosylate the low-molecular-mass GTP-binding proteins of the Rho family, which participate in the regulation of the actin cytoskeleton. ADP-ribosylation inactivates the regulatory Rho proteins and disturbs the organization of the actin cytoskeleton.

Molecular interactions between G-actin, DNase I and the beta-thymosins in apoptosis: a hypothesis

The beta-thymosins are a family of < 5kDa (MW), mostly acidic, proteins which were originally defined in the immune system. Recently, specific members of this family of cytoplasmic polypeptides, namely beta-4 and beta-10, were shown to bind monomeric G-actin both in vitro and in vivo. Whilst many aspects of programmed cell death or 'apoptosis' remain to be defined, the Ca2+/Mg(2+)-dependent endonuclease, DNase I does feature in this process. Monomeric G-actin binds to and inhibits the DNA-degrading activity of DNase I. Given that the intracellular abundance of thymosins beta-4 and beta-10 is related to cell division and differentiation and that anticancer/morphogenic agents such as retinoic acid (RA) and cyclic AMP modulate expression of their respective genes, it is possible that these G-actin sequestering proteins play significant roles in apoptosis perhaps mediated via DNase I.

Interaction of F-actin with phosphate analogues studied by differential scanning calorimetry

The thermal unfolding of F-actin and the changes induced in it by the binding of phosphate analogues were studied by differential scanning calorimetry. It is shown that the conformation of actin is drastically altered by interaction with beryllium fluoride or aluminium fluoride, while the effects of vanadate and phosphate are negligible. The effect of beryllium fluoride on the F-actin structure, as reflected in a significant increase of the actin thermal stability, is much more pronounced in the presence of Mg2+ than in the case of F-actin polymerized by KCl or LiCl in the absence of Mg2+. It is concluded that differential scanning calorimetry is a very convenient method for probing the conformational changes in F-actin caused by the interaction with phosphate analogues.

A yeast actin-related protein homologous to that in vertebrate dynactin complex is important for spindle orientation and nuclear migration

Spindle orientation controls nuclear migration and segregation during mitosis. In yeast, defects in dynein and astral microtubules lead to abnormal spindle orientation and nuclear migration. Dynactin complex is necessary for dynein-mediated vesicle motility in vitro. The major polypeptide of dynactin complex is an actin-related protein in the family Arp1. We have identified in S. cerevisiae a novel actin-related gene, ACT5, in the Arp1 family. An act5 null mutant has defects in spindle orientation and nuclear migration, as does overexpression of Act5p. The phenotype of a double mutant lacking dynein and Act5p is similar to that of single mutants. Therefore, dynactin complex is in the same pathway as dynein and may be necessary for the action of dynein in vivo.

Domain motion in actin observed by fluorescence resonance energy transfer

Actin is composed of two well-separated globular domains which are further subdivided into two subdomains [Kabsch, W., Mannherz, H. G., Suck, D., Pai, E. F., & Holmes, K. C. (1990) Nature 347, 37-44]. Subdomains 1 and 2 constitute the small domain, and subdomains 3 and 4 comprise the large domain. In order to test a hinge bending domain motion in actin such as observed in many kinases, fluorescence resonance energy transfer between two probes attached to each of the two domains was measured by steady-state and time-resolved fluorometers. The adenine base is bound in a hydrophobic pocket between subdomains 3 and 4, and Tyr-69 is located at subdomain 2. In the present study, the adenine moiety was labeled with a fluorescence donor, epsilon ATP, and tyrosine-69 was labeled with the energy acceptor, dansyl chloride. Assuming the random orientation factor k2 = 2/3, the distance between epsilon-adenine moiety and dansyl chloride attached to Tyr-69 in G-actin was determined to be 2.46 nm from steady-state fluorescence measurements. The addition of DNase I did not appreciably change the distance (less than 0.1 nm). The distance decreased to 2.27 nm during polymerization by the addition of phalloidin under physiological salt conditions. On the other hand, time-resolved fluorescence energy transfer measurements have been used to investigate a distribution of distances for a donor-acceptor pair. In G-actin, the mean distance between probes was 2.79 nm with a full width at half-maximum of 3.91 nm, indicating a large number of conformational substates in solution.(ABSTRACT TRUNCATED AT 250 WORDS)

An essential gene of Saccharomyces cerevisiae coding for an actin-related protein

Actin filaments provide the internal scaffold of eukaryotic cells; they are involved in maintenance of cell shape, cytokinesis, organelle movement, and cell motility. The major component of these filaments, actin, is one of the most well-conserved eukaryotic proteins. Recently genes more distantly related to the conventional actins were cloned from several organisms. In the budding yeast, Saccharomyces cerevisiae, one conventional actin gene, ACT1 (coding for the filament actin), and a so-called actin-like gene, ACT2 (of unknown function), have so far been identified. We report here the discovery of a third member of the actin gene family from this organism, which we named ACT3. The latter gene is essential for viability and codes for a putative polypeptide, Act3, of 489 amino acids (M(r) = 54,831). The deduced amino acid sequence of Act3 is less related to conventional actins than is the deduced amino acid sequence of Act2, mainly because of three unique hydrophilic [corrected] segments. These segments are found inserted into a part of the sequence corresponding to a surface loop of the known three-dimensional structure of the actin molecule. According to sequence comparison, the basal core structure of conventional actin may well be conserved in Act3. Our findings demonstrate that, unexpectedly, there exist three members of the diverse actin protein family in budding yeast that obviously provide different essential functions for survival.

Microsecond rotational dynamics of F-actin in ActoS1 filaments during ATP hydrolysis

Rabbit skeletal muscle F-actin labeled at Cys374 with the triplet probe erythrosin-5-iodoacetamide had a steady-state phosphorescence anisotropy (rp) of 0.090 +/- 0.005 at 20 degrees C in 100 mM KCl, pH 7.0, buffer. Titration with skeletal muscle S1 fragment increased rp to 0.138 +/- 0.006 at a mole ratio of 1:1. In the presence of ATP, the anisotropy of the actoS1 complex initially decreased to 0.050 +/- 0.005, a value significantly smaller than the anisotropy of pure F-actin; rp subsequently increased to 0.126 +/- 0.002. The time course of the increase matched that expected from the measured actin-activated ATPase of S1. The plateau value at long time, 0.126, was identical to that of actoS1 in the presence of exogenous ADP or ADP plus phosphate. Characterization of the spectroscopic properties of the erythrosin probe indicated that the changes in rp were not due to changes in fast probe motions on the surface of the filament or the phosphorescence emission lifetime, or in the orientation of the probe on the surface of F-actin, suggesting that they reflect large-scale changes in the microsecond rotational dynamics of actin. ATP hydrolysis by actoS1 thus appeared to induce rotational motions of or within F-actin on the phosphorescence time scale (approximately 300 microseconds). Although the precise physical origin of the induced rotational motions is unknown, this study provides direct evidence that large-scale conformational fluctuations of the actin filament are associated with the force-generating event in actomyosin.

How muscle may contract

A new molecular model is proposed for muscle contraction, that involves the electrical charging of the long (C-terminal) alpha-helical part of the head of the myosin molecule (S1) while the head is attached to actin; as it charges the alpha-helical part moves in the radial electric field between the filaments. The alpha-helical part snaps back when the myosin molecule is discharged electrically, at the moment that ATP binds to the active enzymatic site. This snap-back model explains several puzzling phenomena in contractility, as well as providing a physical explanation for the origin of an impulsive force that drives muscle contraction.

Ultrastructural analysis of the dynactin complex: an actin-related protein is a component of a filament that resembles F-actin

The dynactin complex visualized by deepetch electron microscopy appears as a short filament 37-nm in length, which resembles F-actin, plus a thinner, laterally oriented filament that terminates in two globular heads. The locations of several of the constituent polypeptides were identified on this structure by applying antibodies to decorate the dynactin complex before processing for electron microscopy. Antibodies to the actin-related protein Arp1 (previously referred to as actin-RPV), bound at various sites along the filament, demonstrating that this protein assembles in a polymer similar to conventional actin. Antibodies to the barbed-end actin-binding protein, capping protein, bound to one end of the filament. Thus, an actin-binding protein that binds conventional actin may also bind to Arp1 to regulate its polymerization. Antibodies to the 62-kD component of the dynactin complex also bound to one end of the filament. An antibody that binds the COOH-terminal region of the 160/150-kD dynactin polypeptides bound to the globular domains at the end of the thin lateral filament, suggesting that the dynactin polypeptide comprises at least part of the sidearm structure.

Determination of the alpha-actinin-binding site on actin filaments by cryoelectron microscopy and image analysis

The three-dimensional structure of actin filaments decorated with the actin-binding domain of chick smooth muscle alpha-actinin (alpha A1-2) has been determined to 21-A resolution. The shape and location of alpha A1-2 was determined by subtracting maps of F-actin from the reconstruction of decorated filaments. alpha A1-2 resembles a bell that measures approximately 38 A at its base and extends 42 A from its base to its tip. In decorated filaments, the base of alpha A1-2 is centered about the outer face of subdomain 2 of actin and contacts subdomain 1 of two neighboring monomers along the long-pitch (two-start) helical strands. Using the atomic model of F-actin (Lorenz, M., D. Popp, and K. C. Holmes. 1993. J. Mol. Biol. 234:826-836.), we have been able to test directly the likelihood that specific actin residues, which have been previously identified by others, interact with alpha A1-2. Our results indicate that residues 86-117 and 350-375 comprise distinct binding sites for alpha-actinin on adjacent actin monomers.

Actin mutations that show suppression with fimbrin mutations identify a likely fimbrin-binding site on actin

Actin interacts with a large number of different proteins that modulate its assembly and mediate its functions. One such protein is the yeast actin-binding protein Sac6p, which is homologous to vertebrate fimbrin (Adams, A. E. M., D. Botstein, and D. G. Drubin. 1991. Nature (Lond.). 354:404-408.). Sac6p was originally identified both genetically (Adams, A. E. M., and D. Botstein. 1989. Genetics. 121:675-683.) by dominant, reciprocal suppression of a temperature-sensitive yeast actin mutation (act1-1), as well as biochemically (Drubin, D. G., K. G. Miller, and D. Botstein. 1988. J. Cell Biol. 107: 2551-2561.). To identify the region on actin that interacts with Sac6p, we have analyzed eight different act1 mutations that show suppression with sac6 mutant alleles, and have asked whether (a) these mutations occur in a small defined region on the crystal structure of actin; and (b) the mutant actins are defective in their interaction with Sac6p in vitro. Sequence analysis indicates that all of these mutations change residues that cluster in the small domain of the actin crystal structure, suggesting that this region is an important part of the Sac6p-binding domain. Biochemical analysis reveals defects in the ability of several of the mutant actins to bind Sac6p, and a reduction in Sac6p-induced cross-linking of mutant actin filaments. Together, these observations identify a likely site of interaction of fimbrin on actin.

Mapping actin surfaces required for functional interactions in vivo

An in vivo strategy to identify amino acids of actin required for functional interactions with actin-binding proteins was developed. This approach is based on the assumption that an actin mutation that specifically impairs the interaction with an actin-binding protein will cause a phenotype similar to a null mutation in the gene that encodes the actin-binding protein. 21 actin mutations were analyzed in budding yeast, and specific regions of actin subdomain 1 were implicated in the interaction with fimbrin, an actin filament-bundling protein. Mutations in this actin subdomain were shown to be, like a null allele of the yeast fimbrin gene (SAC6), lethal in combination with null mutations in the ABP1 and SLA2 genes, and viable in combination with a null mutation in the SLA1 gene. Biochemical experiments with act1-120 actin (E99A, E100A) verified a defect in the fimbrin-actin interaction. Genetic interactions between mutant alleles of the yeast actin gene and null alleles of the SAC6, ABP1, SLA1, and SLA2 genes also demonstrated that the effects of the 21 actin mutations are diverse and allowed four out of seven pseudo-wild-type actin alleles to be distinguished from the wild-type gene for the first time, providing evidence for functional redundancy between different surfaces of actin.

Intracellular localization and biochemical function of variant beta-actin, which inhibits metastasis of B16 melanoma

We analyzed the biochemical nature of beta m-actin protein found in mouse B16 melanoma. When we carried out immunostaining with the antibody specific to beta m-actin, filamentous immunofluorescence was observed in B16-F1, a low-metastatic cell line expressing beta m-actin, but not in highly metastatic B16-F10 that did not express beta m-actin. When a purified actin fraction containing beta m-actin was polymerized and immunoprecipitated with anti-beta m-actin antibody, the immunoprecipitate contained beta m-, beta- and gamma-actin. This indicated that the beta m-actin was incorporated into an actin filament together with beta- and gamma-actin in vitro, and this phenomenon was consistently suggested by cellular double immunostaining with anti-beta m-actin and common anti-actin antibody. When the actin fraction containing beta m-actin under a regular depolymerizing condition was subjected to immuno-adsorption assay using anti-beta m antibody and protein-A Sepharose, the immunoadsorbed aggregates contained beta m-, beta- and gamma-actin. This indicates that the actin fraction was not completely depolymerized and contained beta m-actin-containing oligomers, which were too small to be precipitated with anti-beta m-actin antibody alone. The incomplete depolymerization of the beta m-actin-containing fraction was also suggested by the much lower DNase 1 inhibition activity of the beta m-actin-containing fraction than that of beta- and gamma-actin fraction. Furthermore, a DNase 1 binding assay showed that cytoplasmic supernatant prepared from B16-F1 under a low-ionic condition contained less monomeric actin than the cytoplasmic preparation from B16-F10. These results suggested that beta m-actin protein in B16 melanoma probably inhibits the dynamic conversion between the monomeric and polymerized forms of actin, leading to a decrease in cell motility and consequently the suppression of invasiveness and metastasis.

Parser for protein folding units

General patterns of protein structural organization have emerged from studies of hundreds of structures elucidated by X-ray crystallography and nuclear magnetic resonance. Structural units are commonly identified by visual inspection of molecular models using qualitative criteria. Here, we propose an algorithm for identification of structural units by objective, quantitative criteria based on atomic interactions. The underlying physical concept is maximal interactions within each unit and minimal interaction between units (domains). In a simple harmonic approximation, interdomain dynamics is determined by the strength of the interface and the distribution of masses. The most likely domain decomposition involves units with the most correlated motion, or largest interdomain fluctuation time. The decomposition of a convoluted 3-D structure is complicated by the possibility that the chain can cross over several times between units. Grouping the residues by solving an eigenvalue problem for the contact matrix reduces the problem to a one-dimensional search for all reasonable trial bisections. Recursive bisection yields a tree of putative folding units. Simple physical criteria are used to identify units that could exist by themselves. The units so defined closely correspond to crystallographers' notion of structural domains. The results are useful for the analysis of folding principles, for modular protein design and for protein engineering.

Beta thymosins as actin binding peptides

The beta thymosins are a highly conserved family of strongly polar 5 kDa polypeptides that are widely distributed among vertebrate classes; most are now known to bind to monomeric actin and inhibit its polymerization. One beta thymosin, beta four, (T beta 4) is the predominant form in mammalian cells, present at up to 0.5 mM. Many species are known to produce at least two beta thymosin isoforms, in some cases in the same cell. Their expression can be separately regulated. When present outside the cell, the N-terminal tetrapeptide of beta four appears to affect cell cycle regulation; beta thymosins or smaller fragments derived from them may have additional regulatory functions. We suggest that many developmental changes in beta thymosin levels within cells and tissues may be related to changes in G-actin pool size.

Different modes of interaction of two peptide fragments from subdomain 4 of rabbit skeletal muscle actin with actin promoters

Previously, we reported that the 2.6 kDa peptide fragment extending from Arg-177 to Tyr-198 in rabbit skeletal muscle actin bound to actin itself and inhibited its polymerization, while the 9.1 kDa peptide extending from Ser-199 to Tyr-279 in actin did not. The 2.6 kDa segment of actin was reported to contain one of the important actin-actin contacts (Hori, K. and Morita, F. (1992) J. Biochem. 112, 401-408). In this paper, we show additional evidence that the rate of salt-induced increase in the fluorescence of pyrene-labeled actin was decreased in the presence of the 2.6 kDa peptide. Conventional actin filaments were only scarcely observed in the presence of the 2.6 kDa peptide under an electron microscope with a steady state of fluorescence increase. Furthermore, the 2.6 kDa peptide was found to sever F-actin into short filament fragments. The 9.1 kDa peptide, on the other hand, neither inhibited the fluorescence increment of pyrene-actin nor severed actin filaments. However, the 9.1 kDa peptide was found to increase the viscosity and fluorescence intensity of pyrene-G-actin and to form short actin filaments in the absence of salts. Contact sites in the 9.1 kDa segment in actin may have a different mode of interaction with adjacent actin promoters in actin filaments from that of the 2.6 kDa segment.

Functional role of a consensus peptide which is common to alpha-, beta-, and gamma-tubulin, to actin and centractin, to phytochrome A, and to the TCP1 alpha chaperonin protein

The TRiC (TCP1 Ring Complex) chaperonin complex participates in the functional folding of actin, centractin, alpha-, beta-, gamma-tubulin, and phytochrome. Each of the cytoskeletal proteins contain a peptide, RK(A,C,T)F/KRAF, located towards the C-terminus, which is homologous to a TCP1 alpha peptide, while the equivalent phytochrome peptide (RLKAF in certain isoforms) is very similar to the KLRAF peptide of TCP1 alpha. We propose that this TCP1 alpha peptide binds to the nascent polypeptides as they emerge from the ribosome, that this binding restricts the folding pathway, and that the TCP1 alpha peptide is subsequently displaced by the synthesis of the consensus peptide. This hypothesis is strongly supported by the crystallographic structure of actin.

Actophorin preferentially binds monomeric ADP-actin over ATP-bound actin: consequences for cell locomotion

Actophorin from Acanthamoeba castellanii severs actin filaments and sequesters actin monomers. Here we report that actophorin binds ADP-bound monomers with higher affinity than ATP-bound monomers. Actophorin is therefore much less efficient at severing actin filaments in the presence of ADP compared to ATP, particularly taking account of the higher critical concentration in ADP. Monomer binding is also reduced in the presence of 25 mM inorganic phosphate (which is assumed to form ADP.Pi-actin). These findings are discussed in the light of observations on the nucleotide specificity of other monomer binding proteins and related to the role of actin in lamellar protrusion and cell locomotion.

Structural mechanisms for domain movements in proteins

We survey all the known instances of domain movements in proteins for which there is crystallographic evidence for the movement. We explain these domain movements in terms of the repertoire of low-energy conformation changes that are known to occur in proteins. We first describe the basic elements of this repertoire, hinge and shear motions, and then show how the elements of the repertoire can be combined to produce domain movements. We emphasize that the elements used in particular proteins are determined mainly by the structure of the interfaces between the domains.

Recognition of distantly related proteins through energy calculations

A new method to detect remote relationships between protein sequences and known three-dimensional structures based on direct energy calculations and without reliance on statistics has been developed. The likelihood of a residue to occupy a given position on the structural template was represented by an estimate of the stabilization free energy made after explicit prediction of the substituted side chain conformation. The profile matrix derived from these energy values and modified by increasing the residue self-exchange values successfully predicted compatibility of heat-shock protein and globin sequences with the three-dimensional structures of actin and phycocyanin, respectively, from a full protein sequence databank search. The high sensitivity of the method makes it a unique tool for predicting the three-dimensional fold for the rapidly growing number of protein sequences.

Interactions of tensin with actin and identification of its three distinct actin-binding domains

Tensin, a 200-kD phosphoprotein of focal contacts, contains sequence homologies to Src (SH2 domain), and several actin-binding proteins. These features suggest that tensin may link the cell membrane to the cytoskeleton and respond directly to tyrosine kinase signalling pathways. Here we identify three distinct actin-binding domains within tensin. Recombinant tensin purified after overexpression by a baculovirus system binds to actin filaments with Kd = 0.1 microM, cross-links actin filaments at a molar ratio of 1:10 (tensin/actin), and retards actin assembly by barbed end capping with Kd = 20 nM. Tensin fragments were constructed and expressed as fusion proteins to map domains having these activities. Three regions from tensin interact with actin: two regions composed of amino acids 1 to 263 and 263 to 463, cosediment with F-actin but do not alter the kinetics of actin assembly; a region composed of amino acids 888-989, with sequence homology to insertin, retards actin polymerization. A claw-shaped tensin dimer would have six potential actin-binding sites and could embrace the ends of two actin filaments at focal contacts.

The covalent maleimidobenzoyl-actin-myosin head complex. Cross-linking of the 50 kDa heavy chain region to actin subdomain-2

We have identified the region of actin involved in the covalent coupling of maleimidobenzoyl-G-actin to the central 50 kDa segment of the myosin-S-1 heavy chain by analyzing the structure of the maleimidobenzoyl-G-actin-S-1 conjugate using selective proteolytic digestions, amino acid sequence determinations and novel cross-linking reactions between S-1 and different maleimidobenzoyl-G-actin derivatives. The cross-linking is shown to occur only on the stretch of residues 48-67 in actin subdomain-2 with Lys-50, residing on the outer part of the DNase-I-binding loop, as the most likely site of cross-linking. Because the maleimidobenzoyl-F-actin-S-1 complex undergoes the same coupling process, the data provide experimental evidence in favor of the recent model of the rigor F-actin-S-1 complex suggesting a close contact between structural elements of the lower domain of the 50 kDa fragment and the top of actin subdomain-2.

ATP binding in peptide synthetases: determination of contact sites of the adenine moiety by photoaffinity labeling of tyrocidine synthetase 1 with 2-azidoadenosine triphosphate

Characterization of the nucleotide binding domain in peptide synthetases was approached by photoaffinity labeling of tyrocidine synthetase 1 (TY1) with 2-azidoadenosine triphosphate (2-azido-ATP). Exposure of TY1 in the presence of photolabel to irradiation with ultraviolet light resulted in a time-dependent covalent modification of the enzyme with a concomitant loss of catalytic activity. Inactivation was not observed if incubation was performed in the absence of either light or the nucleotide analogue. Specificity of labeling was indicated by the ability of 2-azido-ATP to serve as a substrate in the amino acid activation reaction. The modified protein was subjected to tryptic digestion, and the fragments labeled by the nucleotide analogue were purified by reverse-phase high-performance liquid chromatography. Sequence analysis identified three tryptic peptides corresponding to residues G373-K384, W405-R416, and L483-K494, derived from the N-terminal half of the TY1 sequence. As this region shows similarity to strongly conserved regions in other peptide synthetases and acyl-CoA synthetases, it is considered to be the region catalyzing aminoacyl adenylate formation. The identified sequences appear to define components of the nucleotide binding domain found in close proximity to the adenine ring in ATP. Conservation of primary structure and homology to other carboxyl-activating enzymes of this superfamily, including peptide synthetases, insect luciferases, and acyl-CoA synthetases, is discussed.

The functions and consensus motifs of nine types of peptide segments that form different types of nucleotide-binding sites

From an analysis of current data on 16 protein structures with defined nucleotide-binding sites consensus motifs were determined for the peptide segments that form such nucleotide-binding sites. This was done by using the actual residues shown to contact ligands in the different protein structures, plus an additional 50 sequences for various kinases. Three peptide segments are commonly required to form the binding site for ATP or GTP. Binding motif Kinase-1a is found in almost all sequences examined, and functions in binding the phosphates of the ligand. Variant versions, comparable to Kinase-1a, are found in a subset of proteins and appear to be related to unique functions of those enzymes. Motif Kinase-2 contains the conserved aspartate that coordinates the metal ion on Mg-ATP. Motif Kinase-3 occurs in at least four versions, and functions in binding the purine base or the pentose. Two protein structures show ATP-binding at a separate regulatory site, formed by the motifs Regulatory-1 and Regulatory-2. Structures for adenylate kinase and guanylate kinase show three different sequence motifs that form the binding site for a nucleoside monophosphate (NMP). NMP-1 and NMP-2 bind to the pentose and phosphate of the bound ligand. NMP-1 is found in almost all the kinases that phosphorylate AMP, CMP, GMP, dTMP, or UMP. NMP-3a is found in kinases for AMP, GMP, and UMP, while NMP-3b binds only GMP. For the binding of NTPs, three distinct types of nucleotide-binding fold structures have been described. Each structure is associated with a particular function (e.g. transfer of the gamma-phosphate, or of the adenylate to an acceptor) and also with a particular spatial arrangement of the three Kinase segments evident in the linear sequence for the protein.

Sulfation of guinea pig megakaryocyte and platelet proteins

This study has explored the sulfation of proteins by guinea pig megakaryocytes and platelets and by human platelets. Guinea pig megakaryocytes were incubated in vitro with [35S]sulfate, and the sulfated proteins were separated from proteoglycans by DEAE-Sephacel chromatography and analyzed by SDS-PAGE. The megakaryocytes esterified sulfate to a number of proteins, with the most extensive label migrating at M(r) 42,000, and a second heavily labeled band at M(r) 103,000 in the 0.1 M NaCl DEAE eluate, and 50 and 180 kDa in the 0.23 M NaCl eluate. [35S]-Labeled GPlb alpha was immunoprecipitated from megakaryocyte Triton X-100 extracts. Guinea pig platelet proteins were labeled in vivo by injection of the animals with a single dose of H2(35)SO4. The platelets were activated with thrombin, and cytoskeletal proteins were isolated after treatment of the activated platelets with Triton X-100. About 20% of the platelet macromolecule-associated [35S]sulfate was incorporated into sulfated proteins, which were recovered primarily in the cytoskeleton. The cytoskeleton-associated sulfate radiolabel migrated on SDS-PAGE primarily with actin and additionally with several higher molecular weight proteins. A M(r) 42,000 [35S]-labeled protein was immunoprecipitated by a monoclonal anti-actin antibody, along with molecules of M(r) 160,000 and 180,000 and some higher M(r) material, from the megakaryocytes labeled in vitro with [35S]sulfate. Actin was labeled on 2D isoelectric focusing/SDS-PAGE gels. In addition, there was a very acidic series of heavily [35S]-labeled 42 kDa proteins with about eight components of different isoelectric points with a pattern identical to the M(r) 40,000 cytoskeletal-associated glycoprotein Pltpg40 isolated by Hildreth et al. (1991, Blood 77:121). We hypothesize that sulfation of the cytoskeletal proteins might be involved in cytoskeletal protein interactions and function.

Interaction of skeletal-muscle myosin subfragment-1 with the actin-(338-348) peptide

The data presented here confirm and provide further experimental evidence that rabbit skeletal-muscle myosin subfragment-1 (S-1) binds to the postulated actin-(338-348) hydrophobic segment [Kabsch, Mannherz, Suck, Pai and Holmes (1990) Nature (London) 347, 37-44] with high affinity in the absence and presence of MgATP. The apparent dissociation constant of the S-1 interaction (5.5 x 10(-7) M) with the actin-(338-348) peptide was of the same order of magnitude as that of the actin-(18-28) binding site (2 x 10(-6) M). In similar conditions, fragmented (27 kDa-50 kDa-20 kDa) S-1 also bound to the peptide. Antibodies directed to the vicinal sequence 348-358 were rapidly eliminated from actin by S-1 interaction and weakened S-1 binding to monomeric or filamentous actin. The antigenic site (348-358) is located very close to the C-terminal S-1-binding site (360-369) and encompasses some residues (Leu-349 and Phe-352) included in the hydrophobic S-1-binding region [Schröder, Manstein, Jahn, Holden, Rayment, Holmes and Spudich (1993) Nature (London) 364, 171-174]. It was observed that anti-[actin-(348-358)] antibodies were also unable to decrease actomyosin ATPase activity, in contrast with previous results obtained with anti-[actin-(18-28)] antibodies [Adams and Reisler (1993) Biochemistry 32, 5051-5056]. The hydrophobic actin-(338-348) peptide used in considerable excess was unable to perturb acto-S-1 and S-1 activities in contrast with results obtained with the N-terminal actin peptide [Kôgler, Moir, Trayer and Ruegg (1991) FEBS Lett. 294, 31-34].

Molecular movements in contracting muscle: towards "muscle--the movie"

The recent publication of the crystal structures of G-actin and of myosin subfragment-1, together with analysis of a time-resolved series of well sampled low-angle 2D X-ray diffraction patterns from bony fish muscle permits the study of the molecular movements in muscle that are associated with generation and regulation of contractile force. Here it is shown that even though low-angle (i.e. low resolution) X-ray diffraction patterns are being used, these patterns are sensitive, for example, to sub-domain movements of as little as 3 A or 4 degrees within the actin monomers of actin filaments. Actin filament diffraction patterns from whole muscle are being used to define actin domain and tropomyosin movements involved in regulation. Myosin and actin filament diffraction patterns are being used together to start to show how the complete "quasi-crystalline" unit cell in the bony fish muscle A-band can be modelled as a series of time-slices through a typical tetanic contraction of the muscle. In this way, the time sequence of images can be used to create "muscle--the movie".

Cysteine-specific ADP-ribosylation of actin

Incubation of lysate from human polymorphonucleated neutrophils and human platelets with [32P]NAD resulted in the labeling of a 42-kDa protein. Phosphodiesterase (Crotalus durissus) released 5'-AMP from the radiolabeled protein. The 42-kDa protein was identified as actin by binding to DNAse-I, two-dimensional gel electrophoresis and partial proteolysis. The rate of ADP-ribosylation was greater with [32P]ADP-ribose than with [32P]NAD, indicating a non-enzymic modification. ADP-ribose also modified actin in the actin-DNAase-I complex, but denatured actin was not modified by ADP-ribose. Only cytoplasmic beta/gamma-actin isoforms were non-enzymically ADP-ribosylated but not muscle alpha-actin. The acceptor amino acid was identified as a cysteine residue whereas the bacterial ADP-ribosyltransferase C. perfringens iota toxin catalyzes incorporation of ADP-ribose to Arg177 of actin. Alkylation of cysteine residues of actin with N-ethylmaleimide prevented subsequent non-enzymic ADP-ribosylation but not the toxin catalyzed modification. Non-enzymically ADP-ribosylated actin was further modified by C. perfringens iota toxin. The F-actin stabilizing mycotoxin phalloidin blocked the non-enzymatic ADP-ribosylation and, conversely, ADP-ribosylation inhibited the phalloidin-induced polymerization of ADP-ribosylated actin. The data indicate that cytoplasmic actin is non-enzymically ADP-ribosylated by ADP-ribose at a cysteine residue to inhibit actin polymerization.

The glycine-rich sequence of protein kinases: a multifunctional element

Evolution favours the use of glycine-rich loops for nucleotide binding in proteins. In the large family of protein kinases, the catalytic domain of which has one of the highest degrees of conservation among all known proteins, the structure of the nucleotide-binding site differs from classical folds. We are now beginning to understand the multiple functional roles of the glycine-rich sequence in protein kinases and some of the structural constraints leading to its conservation.

Mapping of the actomyosin interfaces

Recombinant DNA methods were used to obtain soluble, undenatured fragments of the heavy chain of myosin subfragment 1 (S-1). These fragments were of preselected lengths and could include protease-sensitive segments that are destroyed when other preparation methods are used. Actin binding by each of the three contiguous segments (residues 1-248, 249-524, and 518-722, essentially spanning the entire S-1 heavy chain) was demonstrated. ATP binding, comparable to that of native S-1, was obtained only with a segment consisting of residues 1-524. Competition among the various fragments for actin was also studied. The data are discussed in relation to the recently reported resolved structure of S-1 [Rayment, I., Rypnieski, R. W., Schmidt-Bäse, K., Smith, R., Tomchick, D. R., Benning, M. M., Winkelmann, D. A., Wesenberg, G. & Holden, H. M. (1993) Science 261, 50-58].

Structure of the regulatory domain of scallop myosin at 2.8 A resolution

The regulatory domain of scallop myosin is a three-chain protein complex that switches on this motor in response to Ca2+ binding. This domain has been crystallized and the structure solved to 2.8 A resolution. Side-chain interactions link the two light chains in tandem to adjacent segments of the heavy chain bearing the IQ-sequence motif. The Ca(2+)-binding site is a novel EF-hand motif on the essential light chain and is stabilized by linkages involving the heavy chain and both light chains, accounting for the requirement of all three chains for Ca2+ binding and regulation in the intact myosin molecule.

The three-dimensional structure of a molecular motor

Myosin is one of only three proteins known to convert chemical energy into mechanical work. Although the chemical, kinetic and physiological characteristics of this protein have been studied extensively, it has been difficult to define its molecular basis of movement. With the recent X-ray structural determination of the myosin head, however, it is now possible to put forward a hypothesis on how myosin might function as a molecular motor.

Cloning and sequencing of the Leishmania major actin-encoding gene

We report the cloning and characterization of the genomic sequence of the actin (Act)-encoding gene (act) from Leishmania major. Restriction maps of two genomic clones, as well as genomic Southern analysis strongly suggest that the act of L. major is a single-copy gene. A single 1.6-kb transcript is detected in Northern blots. The deduced amino-acid sequence shows 68-89% identity with Act sequences from other eukaryotes.

Three-dimensional reconstruction of a co-complex of F-actin with antibody Fab fragments to actin's NH2 terminus

We have decorated F-actin with Fab fragments of antibodies to actin residues 1-7. These antibody fragments do not strongly affect the rigor binding of myosin S-1 to actin, but do affect the binding of S-1 to actin in the presence of nucleotide (DasGupta, G., and E. Reisler, 1989. J. Mol. Biol. 207:833-836; 1991. Biochemistry. 30:9961-9966; 1992. Biochemistry. 31:1836-1841). Although the binding constant is rather low, we estimate that we have achieved about 85% occupancy of the actin sites. Three-dimensional reconstructions from electron micrographs of both negatively stained and frozen-hydrated filaments show that the Fab fragment is bound at the location of the NH2 terminus in the model of Holmes et al. (Holmes, K.C., D. Popp, W. Gebhard, and W. Kabsch. 1990. Nature. 347:37-44) for F-actin, excluding very different orientations of the actin subunit in the filament. Most of the mass of the antibody is not visualized, which is due to the large mobility of the NH2 terminus in F-actin, differences in binding angle within the polyclonal antibody population, or a combination of both of these possibilities.

Structural aspects of actin-binding proteins

The three-dimensional structures of myosin subfragment 1 (S1), gelsolin segment 1 complexed with alpha-actin, villin fragment 14T, Acanthamoeba profilin-I, and bovine profilin complexed with beta-actin were completed last year. Together, they expand our understanding of the structural organization of actin-binding proteins. In addition, the segment 1 and bovine profilin complexes provide atomic-level descriptions of their interfaces with actin.

Three-dimensional structure of a single filament in the Limulus acrosomal bundle: scruin binds to homologous helix-loop-beta motifs in actin

Frozen, hydrated acrosomal bundles from Limulus sperm were imaged with a 400 kV electron cryomicroscope. Segments of this long bundle can be studied as a P1 crystal with a unit cell containing an acrosomal filament with 28 actin and 28 scruin molecules in 13 helical turns. A novel computational procedure was developed to extract single columns of superimposed acrosomal filaments from the distinctive crystallographic view. Helical reconstruction was used to generate a three-dimensional structure of this computationally isolated acrosomal filament. The scruin molecule is organized into two domains which contact two actin subunits in different strands of the same actin filament. A correlation of Holmes' actin filament model to the density in our acrosomal filament map shows that actin subdomains 1, 2, and 3 match the model density closely. However, actin subdomain 4 matches rather poorly, suggesting that interactions with scruin may have altered actin conformation. Scruin makes extensive interactions with helix-loop-beta motifs in subdomain 3 of one actin subunit and in subdomain 1 of a consecutive actin subunit along the genetic filament helix. These two actin subdomains are structurally homologous and are closely spaced along the actin filament. Our model suggests that scruin, which is derived from a tandemly duplicated gene, has evolved to bind structurally homologous but non-identical positions across two consecutive actin subunits.

Mechanisms of leukocyte motility and chemotaxis

Motility is a complex process that depends on the coordination of many cellular functions, including the conversion of information from the environment into a series of coordinated responses that culminate in directed cell movement. Major advances have been made in the understanding of many functions involved in motility, such as transmembrane signaling events, leading to alterations in the actin cytoskeleton, and interactions between adhesion receptors and components of the cytoskeleton, providing a link between the extracellular and intracellular environments. Studies using yeast (Saccharomyces cerevisiae), slime molds (Dictyostelium discoideum) and nematodes (Caenorhabditis elegans) have advanced our understanding of the molecular biology of cytoskeletal proteins and have important implications for mammalian leukocyte motility.

Effect of phalloidin on the ATPase activity of striated muscle myofibrils

Phalloidin was shown to increase the ATPase activity and Ca2+ sensitivity of both bovine cardiac and rabbit psoas myofibrils when assayed in a solution containing 50 mM KCl, 100 mM MOPS (pH 7.0), 2 mM MgCl2, 1 mM ATP, 2 mM EGTA, and varying concentrations of Ca2+ (temperature 21-22 degrees C). The phalloidin effect in cardiac myofibrils developed over a time course of several minutes in the presence of 50 microM phalloidin. Relative increase of ATPase activity was maximal at pCa 8 and decreased with decrease in pCa. In cardiac myofibrils the increase was about 70% at pCa 8 and 20% at pCa 4 following 20-30 min pre-incubation with 2 microM or 50 microM phalloidin. The effect persisted after excess phalloidin was washed out. The increase in Ca2+ sensitivity was approximately 0.15 pCa units. For skeletal myofibrils treated with 2 microM phalloidin all changes were considerably less than those seen with cardiac myofibrils and the changes were even less when the myofibrils were exposed to 50 microM phalloidin. These results show that when specifically bound to actin, phalloidin can change the kinetic parameters of the cross-bridge cycle and may also alter the Ca2+ sensitivity of the contractile system. The effects of phalloidin seem to vary with muscle type.