Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M601973200 on August 18, 2006

J. Biol. Chem., Vol. 281, Issue 42, 31909-31919, October 20, 2006
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
281/42/31909    most recent
M601973200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rould, M. A.
Right arrow Articles by Trybus, K. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rould, M. A.
Right arrow Articles by Trybus, K. M.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Crystal Structures of Expressed Non-polymerizable Monomeric Actin in the ADP and ATP States*

Mark A. Rould, Qun Wan, Peteranne B. Joel, Susan Lowey, and Kathleen M. Trybus1

From the Department of Molecular Physiology and Biophysics, University of Vermont, Burlington, Vermont 05405

Received for publication, March 1, 2006 , and in revised form, August 14, 2006.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Actin filament growth and disassembly, as well as affinity for actin-binding proteins, is mediated by the nucleotide-bound state of the component actin monomers. The structural differences between ATP-actin and ADP-actin, however, remain controversial. We expressed a cytoplasmic actin in Sf9 cells, which was rendered non-polymerizable by virtue of two point mutations in subdomain 4 (A204E/P243K). This homogeneous monomer, called AP-actin, was crystallized in the absence of toxins, binding proteins, or chemical modification, with ATP or ADP at the active site. The two surface mutations do not perturb the structure. Significant differences between the two states are confined to the active site region and sensor loop. The active site cleft remains closed in both states. Minor structural shifts propagate from the active site toward subdomain 2, but dissipate before reaching the DNase binding loop (D-loop), which remains disordered in both the ADP and ATP states. This result contrasts with previous structures of actin made monomeric by modification with tetramethylrhodamine, which show formation of an {alpha}-helix at the distal end of the D-loop in the ADP-bound but not the ATP-bound form (Otterbein, L. R., Graceffa, P., and Dominguez, R. (2001) Science 293, 708-711). Our reanalysis of the TMR-modified actin structures suggests that the nucleotide-dependent formation of the D-loop helix may result from signal propagation through crystal packing interactions. Whereas the observed nucleotide-dependent changes in the structure present significantly different surfaces on the exterior of the actin monomer, current models of the actin filament lack any actin-actin interactions that involve the region of these key structural changes.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
ATP hydrolysis drives actin filament dynamics. The effect known as treadmilling arises from the preferential addition of ATP-actin monomers to the plus or barbed end of actin and the preferential dissociation of ADP-actin from the minus or pointed end. Hydrolysis of ATP occurs after the monomer is incorporated into the filament. In addition, actin-binding proteins often have significantly different affinities for the ADP-bound versus ATP-bound forms of actin. Both of these lines of evidence strongly suggest that there must be significant structural changes in actin induced by ATP hydrolysis. Indirect solution techniques such as proteolytic digestion rates and fluorescence studies (reviewed in Ref. 1) are in agreement with conformational differences between the ADP and ATP states, but direct structural evidence has been lacking until recently.

Based on crystal structures of a tetramethylrhodamine-labeled monomeric actin (TMR-actin)2 with ADP (2) or AMP-PNP (3) at the active site, Dominguez and co-workers proposed the provocative idea that ATP hydrolysis initiates a series of changes originating at the active site, that ultimately cause a loop-to-helix transition in the DNase binding loop in subdomain 2 of actin ("D-loop," residues in subdomain 2 of actin which comprise part of the DNase I binding site). They suggested that this was the long sought after change in structure between ADP and ATP actin. The cleft between subdomains 2 and 4 of actin remained closed in both nucleotide states. Despite this observation, an opposing point of view (4) held that the more important conformational change is an opening of the cleft upon ATP hydrolysis, a change that would be compatible with that observed for other nucleotide hydrolyzing proteins. Docking of actin crystal structures into negatively stained images of F-actin was also consistent with the idea that ADP-actin had a more open cleft than the triphosphate state (5). If the latter view is correct, one would have to suppose that the modification of Cys374 by TMR stabilized the closed conformation and inactivated the nucleotide sensing mechanism by virtue of the binding of rhodamine between subdomains 1 and 3, a potential hinge region (6) of the molecule. It has been suggested (4) that the short helix observed in the crystal structure of the ADP state of TMR-actin could be formed as a result of contacts with neighboring molecules in the crystal. We provide further evidence to support this idea, and suggest a pathway through which nucleotide-dependent changes may propagate through fortuitous crystal packing interactions from one monomer to the D-loop region of another in the TMR-actin crystals.

Here we crystallized and expressed a cytoplasmic actin that was rendered incapable of polymerization by virtue of two surface mutations in subdomain 4 (A204E/P243K). This strategy negates concerns raised about the TMR modification of actin. De novo crystallization of AP-actin with either ATP or ADP at the active site reveals obligatory nucleotide-dependent conformational changes that are localized to the active site and the so-called "sensor loop" (residues 71-73 in actin, Ref. 3), and propagate only short distances from there. A helix is not observed in the D-loop in either the ADP- or ATP-bound states, and the cleft remains closed in both states. Our data show that a change in the D-loop conformation and/or a change in cleft disposition is not an obligatory nucleotide-dependent change in the actin monomer.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Expression and Purification of AP-actin—A modified plasmid pAcUW2B containing the coding sequence for the Drosophila 5C actin gene was a generous gift from Volkman et al. (7). Site-directed mutagenesis was used to create a mutant construct of the 5C actin in which Ala204 was replaced by Glu and Pro243 was replaced by Lys (AP-actin). The numbering system used here corresponds to that of skeletal muscle actin. 5C actin is a cytoplasmic actin, and thus it has one less acidic residue at the N terminus compared with striated muscle actins, which have four. Drosophila 5C actin is 98.7% identical with human {gamma}-cytoplasmic actin.

Infection of Sf9 cells with recombinant baculovirus encoding the AP-actin construct, lysis of the cells, and dialysis of the cell lysate have been described (8). The dialysate was fractionated on a Q-Sepharose column (1.5 x 30 cm for a 4 billion cell culture preparation) using a 500-ml gradient from 0.1 to 0.4 M KCl (10 mM imidazole, pH 7.5, 0.1 mM CaCl2, 0.5 mM DTT, 0.5 mM Na2ATP, 1 µg/ml leupeptin). Pooled fractions from the Q Sepharose were concentrated with an Amicon Ultacentrifugal filter device (Millipore Corp.) and applied to a Sephacryl S300 column as described (8). Some preparations were run on a second Q-Sepharose column following the Sephacryl S300 column. The final preparation of AP-actin was dialyzed into G buffer (5 mM Tris-HCl, pH 8.2 at 4 °C, 0.2 mM CaCl2, 0.1 mM sodium azide, 0.5 mM DTT, 1 µg/ml leupeptin, and 0.2 mM Na2ATP or Na2ADP and stored in liquid nitrogen. AP-actin concentration was determined from the absorbance at 290 nm by use of an extinction coefficient of 0.63 ml mg-1 cm-1. Chicken skeletal muscle actin was prepared as described in Ref. 9.

Conversion of ATP-actin to ADP-actin—To prepare AP-actin with ADP in the active site, AP-actin dialyzed into G buffer with Na2ADP was diluted to 9 or 12 mg/ml with G Buffer. ADP was increased to 0.5 mM. 4 mM MgCl2 and 0.5 mM EGTA were added, and the solution was incubated 10 min on ice. 2 mM glucose and 40 units/ml hexokinase (Sigma) were added, and the solution was incubated on ice for at least 40 h (10). The solution was then centrifuged 20 min at 300,000 x g, the concentration was adjusted to 7 or 10 mg/ml, and 5 mM DTT was added. To prepare AP-actin with ATP in the active site, AP-actin dialyzed into G buffer with Na2ATP was treated the same as the AP-actin-ADP except that ATP rather than ADP was added, and no hexokinase was used.

Nucleotide Hydrolysis—Free external nucleotides were removed from CaATP-AP-actin in G buffer by addition of 10% by volume of a 50% slurry of Dowex AG-1 x 8. The suspension was mixed 1 min on ice and centrifuged 20 s at 10,000 x g. The Dowex treatment was repeated two more times. The final supernatant was centrifuged 20 min at 300,000 x g. The CaATP-AP-actin was adjusted to 4 mg/ml. A portion of this protein was converted to MgATP-AP-actin by addition of 0.2 mM MgCl2 and 0.5 mM EGTA. To determine the rate of nucleotide hydrolysis, samples were incubated on ice, and 50-µl samples were taken at various time intervals. To liberate nucleotide from the active site of actin and precipitate the protein, 50 µlof 10% perchloric acid was added to the sample. After 5 min on ice, 20 µl of 4 M potassium acetate in 10 M KOH was added to neutralize the solution, and the sample was centrifuged for 10 min at 12,000 x g followed by 10 min at 300,000 x g. The supernatant was diluted to 200 µl with water and loaded on a 1-ml Mono Q HR5/5 column (Amersham Biosciences) equilibrated with 5 mM triethylammonium bicarbonate pH 8.5 (Sigma-Aldrich) using an AKTA-FPLC system (GE Healthcare). The nucleotides were eluted with a 10-ml gradient of 5 mM to 0.5 M triethylammonium bicarbonate pH 8.5, followed by an additional 4 ml of 0.5 M triethylammonium bicarbonate pH 8.5. ATP and ADP standards were 25 µM.

Nucleotide Exchange—The rate of nucleotide exchange in AP-actin and in tissue-purified skeletal G-actin was determined by measuring the decrease in fluorescence on release of etheno-ATP ({epsilon}-ATP) (Molecular Probes) from actin. Unbound ATP was removed from 5 µM actin by addition of 10% by volume of 50% Dowex AG-1 x 8 followed by centrifugation to remove the resin. 100 µM {epsilon}-ATP was then added to the actin solution, and the solution was left on ice for 2 h to allow {epsilon}-ATP to exchange into the active site. The actin solution was treated with Dowex AG-1 x 8 to remove unbound nucleotide. The fluorescence of bound {epsilon}-ATP (2 µM actin, 5 mM Hepes, pH 7.5, 50 µM CaCl2, 1 mM DTT, 20 °C) was measured in a fluorometer (model K2, ISS, Inc; 340 nm excitation, 415 nm emission). 100 µM Na2ATP was added, and the decrease in fluorescence as ATP was exchanged for {epsilon}-ATP was followed as a function of time. Data were fitted to a single exponential.

Subtilisin Digestion—Skeletal G-actin and AP-actin (10 µM) were cleaved by subtilisin at an enzyme/protein mass ratio of 1:1500 at 25 °C, as described in Ref. 11. ATP-actin and ADP-actin were prepared as described in a previous section.

Crystallization—Crystals of AP-actin (10-11 mg/ml in G buffer) were first obtained by vapor diffusion at 4 °C. The protein was mixed with an equal volume of reservoir buffer composed of 25% 2-methyl-2,4-pentanediol, 100 mM NaAc, pH 4.8, 100 mM NaCl, 40 mM CaCl2. Optimal crystals were obtained by several rounds of microseeding. Preparation of ADP-actin or ATP-actin for microseeding is described above. The composition of the reservoir buffer for the microseeding experiments was 20% 2-methyl-2,4-pentanediol, 100 mM NaAc, pH 5.4, 100 mM NaCl, 20 mM CaCl2 for ADP-actin crystals, and the same solution but at pH 5.2 for the ATP-actin crystals. Crystals were flash frozen in liquid nitrogen directly from the crystallization buffer.

Structure Determination and Refinement—Diffraction data were collected on a Rigaku RUH3R generator and MAR 345 detector at a temperature of 100K maintained by a CryoIndustries cryocooler. Crystals of ADP- and ATP-bound forms of AP-actin belong to space group C2, with cell parameters a = 199.7 Å, b = 54.07 Å, c = 39.59 Å, and beta = 93.16°. Data collection statistics are provided in Table 1. The ATP-bound diffraction data and refinement statistics are of higher quality despite a higher overall B because these data were collected using Xenocs multilayer optics, while data for the ADP-bound crystal were collected using double mirror optics.


View this table:
[in this window]
[in a new window]
  		TABLE 1
Data collection and model refinement statistics

 
The structure of ADP-bound AP-actin was solved by molecular replacement via EPMR (12) using only the protein portion of TMR-Actin-ADP (PDB: 1J6Z [PDB] ) as the search model. All atomic B-factors were reset to 10 Å2 and the model was rebuilt into simulated-annealing Fo-Fc omit maps (T0 = 9000 K) (13) after rigid-body refinement. Cycles of rebuilding into SA-omit maps using the molecular graphics program vuSette zc (M.A.R.) and refinement with CNS (14) yielded a final model with statistics given in Table 1.

A model of the ADP state above, at an early stage in refinement, with all water molecules and ligands removed and all B-factors reset, served as the initial model for building the ATP-bound state of AP-actin. The model was rebuilt into simulated-annealing omit maps (T0 = 9000 K) and refined as above, independently of the ATP-bound structure. The same set of reflections (about 10% of the total) were set aside from both the ATP- and ADP-actin data sets for cross-validation (i.e. for the "free R-factor" calculation (15)), prior to any refinement.

Because the ATP- and ADP-bound AP-actin crystals are isomorphous (i.e. belong to the same space group, have the same unit cell parameters, and show a close agreement in diffraction amplitudes), we are able to make use of isomorphous differential crystallography to accurately visualize the differences between the two structures. These methods allow a direct comparison of the structures at the level of differences in their electron density, free of model bias (16). Given that the root mean square deviation (RMSD) after superposition of the two {alpha}-carbon backbones is only 0.23 Å and the estimated model coordinate error is also 0.23 Å (Table 1), differential crystallography is essential for the detection of any subtle yet significant propagated structural changes arising from the change in nucleotide state. To generate an isomorphous difference electron density map, the scaled differences between the observed diffraction amplitudes from the ATP- and ADP-bound actin crystals are used as coefficients in a Fourier transform, along with phases calculated from either of the refined models. Positive density (shown as green contours, as in Fig. 4) in the resulting isomorphous difference Fourier map indicates regions where the density in the ATP state is greater than in the ADP state, and vice versa for negative density (red contours). Peaks of positive difference density adjacent to peaks of negative difference density indicate that an atom or group has shifted, with a detection limit of less than a tenth of an angstrom. Further details of the procedure are in Ref. 17.

The isomorphous character of the AP-actin-ATP and -ADP crystals allows for an additional means of validating the small coordinate shifts seen between the independently refined models of the two structures. After refinement is complete, the two models and their corresponding diffraction data are swapped and re-refined; that is, a new "swapped" AP-actin-ATP model is refined starting from the final refined AP-actin-ADP coordinates, and vice versa. Crystallographic refinement of atomic positions and individual B-factors via CNS (14) is applied to convergence for the swapped models of both states, using the corresponding diffraction amplitudes (i.e. the new swapped ATP-state model is the original ADP-state model refined against the ATP-state diffraction data). Validation of the differences between the ATP- and ADP-bound crystal structures of actin is determined by the degree to which the atomic shift vectors between refined swapped ATP- and ADP-bound models recapitulate the atomic shift vectors between the original (final refined unswapped) models. Quantitatively this is assessed as the vector correlation between ATP-to-ADP atomic shift vectors for the original and swapped models, summed over all {alpha}-carbons in Equation 1,

Formula 1(Eq. 1)
where the vector O is the coordinate difference of an atom between the original refined ADP-bound and ATP-bound crystal structures, and S is the same coordinate difference after swapping models and re-refining. Intuitively, the dot product O·S represents the extent to which the two vectors point in the same direction, weighted by the product of the magnitude of the two vectors. This swapped-refined-shift-vector-correlation statistic is analogous to the common scalar correlation coefficient and thus has the range of -1 (perfect anti-correlation) to +1 (perfect correlation.) For the ATP- and ADP-bound AP-actin models reported herein, the swapped-refined-shift-vector-correlation is 0.80 (excluding residues 70-73 of the sensor loop, whose shifts are outside the radius of convergence of crystallographic refinement without manual re-building), suggesting that the modeled coordinate differences between the states represent well the true differences, however subtle, present in the crystals. Further details of the procedure are in Ref. 17.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Biochemical Properties of AP-actin—Two surface mutations in subdomain 4 of actin (A204E/P243K) are sufficient to render actin completely non-polymerizable (8). As previously described (8), these mutations were chosen based on their presence in the sequence of a non-polymerizable actin-related protein, Arp3 (18). The mutant actin, named AP-actin for the residues present in the wild-type structure, was expressed in high yield (≥ 60 mg/liter culture) in the baculovirus/insect cell expression system. The actin isoform used here is 98.7% identical with human {gamma}-cytoplasmic actin. Our approach to crystallizing this protein is unique in that it is the first monomeric actin crystallized without chemical cross-linking, toxins, or actin-binding proteins to prevent polymerization.

Because AP-actin shows no tendency to polymerize, ATP hydrolysis is very slow. Rapid hydrolysis of actin-bound ATP only occurs after a monomer has added onto the filament (reviewed in (19)). Analysis of the nucleotide content of AP-actin with Ca2+ as the divalent cation in the active site showed little hydrolysis of CaATP over the course of 72 h. With Mg2+ as the divalent cation, ATP hydrolysis occurred at a rate of 0.025 h-1 (Fig. 1A). This result established that AP-actin is catalytically active, despite being polymerization-incompetent. One proposed explanation for the reduced tendency of CaATP to be hydrolyzed is that the bond distance and angle between the {gamma}-phosphate of ATP and a key water molecule that is poised to act as a nucleophile is more optimally aligned when magnesium is coordinated than when calcium is the divalent cation (20). In contrast, it was necessary to use the non-hydrolyzable analog AMPPNP to crystallize TMR-actin in the nucleotide triphosphate state, presumably because TMR modification does not render the molecule as non-polymerizable as the two surface mutations in AP-actin (21).


Figure 1
View larger version (14K):
[in this window]
[in a new window]
  		FIGURE 1.
Properties of AP-actin in solution. A, ATP hydrolysis by CaATP-AP actin (open circles) and MgATP-AP-actin (filled circles) as a function of time. The data obtained in magnesium were fitted to a single exponential equation with a rate of 0.025 h-1. ATP hydrolysis in CaATP-AP-actin was negligible. Data are from two experiments with two different preparations of AP-actin. B, rate of {epsilon}-ATP release from tissue-purified skeletal G-actin and AP-actin. The normalized rate of {epsilon}-ATP release from G-actin (filled circles) and AP-actin (open circles) were fitted to a single exponential equation with rates of 0.006 s-1 and 0.010 s-1, respectively. C, subtilisin cleavage of skeletal G-actin and AP-actin containing either ATP or ADP. The inset shows SDS gels of the time course of AP-ADP and AP-ATP digestion by subtilisin. The upper bands are intact actin, and the lower bands, the 35-kDa-digested fragment. The SDS-PAGE patterns were quantified, and the percent of undigested actin was plotted versus time on a semilogarithmic plot. Data were fitted to a single exponential equation. With CaATP bound, the rates of cleavage for skeletal G-actin (filled circles) and AP-actin (open circles) were 0.135 and 0.114 min-1, respectively. With MgADP bound, the rates of cleavage for skeletal G-actin (filled triangles) and AP-actin (open triangles) were 0.024 and 0.014 min-1, respectively.

 
The rate of ATP release from the active site of AP-actin is very similar to that of tissue-purified skeletal G-actin, implying that the two point mutations have little effect on this fundamental property of actin (Fig. 1B). Subtilisin digestion between Met47 and Gly48 in subdomain 2 was also used to show that the point mutations had no long range effect on other domains of actin. The time course of the digestion pattern of AP-actin and tissue-purified skeletal actin with ATP at the active site was essentially indistinguishable (Fig. 1C, circles). Moreover, the rate of digestion of both AP-actin and tissue-purified skeletal actin was considerably slower in the ADP than in the ATP state (Fig. 1C, triangles). The slower digestion in ADP has been considered to be evidence that the D-loop undergoes a nucleotide-dependent change in conformation (22).

We also tested whether cofilin would bind more tightly to the ADP than the ATP form of AP-actin, as has been established for skeletal muscle actin (23). Gel filtration chromatography in G buffer with 0.2 M NaCl was used to show that ADP-AP-actin forms a stoichiometric complex with cofilin. Under the same conditions, no complex formation was observed with ATP-AP-actin (data not shown). These biochemical assays established that AP-actin retains many important properties of skeletal muscle G-actin, except for its ability to remain monomeric at high concentrations, which was the intended goal of the mutations.

Structure of AP-actin with ATP or ADP at the Active Site— AP-actin was first crystallized with CaATP at the active site, the nucleotide that remained bound throughout the purification procedure. Calcium was also a component of the solution used for crystallization. The ATP-actin crystal diffracted to 1.8 Å (see Table 1). An overview of the structure, with the location of the point mutations indicated, is shown in Fig. 2A. The overall structure is strikingly similar to most previously published structures (reviewed in Ref. 4); the RMS residual after superimposing the {alpha}-carbon backbones of AP-actin with TMR-actin (with AMPPNP bound, PDB code 1NWK) at the active site (Fig. 3B) is only 1.06 Å. The nucleotide-binding cleft is closed, and the D-loop in subdomain 2 is disordered. Importantly, the mutated residues caused no global or local changes in the structure (Fig. 2B).

ADP was exchanged for ATP at the active site, and ADP-actin was crystallized under conditions that were only subtly different from those used for ATP-actin (see "Materials and Methods"). The most striking feature of the ADP-actin is its remarkable overall similarity to the ATP form (Fig. 3A). The cleft remains closed and the D-loop in subdomain 2 remains disordered. The short {alpha}-helix in the D-loop seen in the ADP form of TMR-actin (2) is not observed for AP-actin (Fig. 3C), even though there is sufficient space for the helix in the AP-actin crystals (discussed below).

Structural Differences between ATP- and ADP-bound AP-Actin—The {alpha}-carbon backbones of the two states of AP-actin superimpose with an RMSD of 0.23 Å, which drops to 0.19 Å excluding residues 70-73 of the sensor loop. All protein atoms between the two structures superimpose with an RMSD of 0.53 Å. The only structural changes in the backbone of actin as a function of bound nucleotide are localized to the active site region (Fig. 3A). This is best illustrated by an isomorphous difference Fourier map (Fobs,ATP-Fobs,ADP) covering the entire actin molecule (Fig. 4A), which shows the differences in unbiased electron density between the two states. Although the map covers the entire actin monomer, only the region of the active site, including the sensor loop, shows any significant change in electron density.


Figure 2
View larger version (32K):
[in this window]
[in a new window]
  		FIGURE 2.
A, schematic rendition of the AP-actin-ATP complex, with domain numbers encircled. The two amino acid changes in domain 4 (A204E and P243K), that render the molecule non-polymerizable, are indicated. Residues 40-49, constituting the DNase I binding loop, are not ordered in either the ADP- or ATP-bound states of AP-actin. (Helices are in cyan, beta-sheets in magenta, the calcium ion is green; side chains of Glu204 and Lys243 are dark blue; and for the nucleotide, carbon is yellow, nitrogen blue, oxygen red.) B, two mutations do not significantly alter the structure, as shown by super-position with the corresponding region of TMR-modified actin (1J6Z, red). The two new mutated residues, Glu204 and Lys243, do not interact.

 
Propagated Structural Changes Accompanying the ADP to ATP Transition—A close-up of the active site region (Fig. 4B) reveals the detailed structural differences between the two states, and readily suggests a mechanism by which the presence of the {gamma}-phosphate of ATP initiates a series of structural transitions extending from the active site outward to the sensor loop at the surface of the protein.

Starting from the actin monomer in its ADP-bound conformation, introduction of the {gamma}-phosphate of ATP immediately impacts the structure. Ser14, which bonds favorably to the beta-phosphate in the ADP state (Figs. 4B and 5), undergoes a severe steric clash with the {gamma}-phosphate of ATP, with the beta-carbon and hydroxyl of Ser14 only 2.4 Å away from the phosphate oxygens. Because the ATP is held tightly in place by a network of bonds and Van der Waals contacts (Fig. 5), Ser14 is obliged to shift 0.6 Å away from the {gamma}-phosphate, and its hydroxyl side chain rotates 130°. The displaced serine in turn impinges on backbone carbonyls of the "sensor loop" (residues 71-73), with the serine {alpha}- and beta-carbons now only 2.9 Å and 2.7 Å away from the carbonyls of Ile71 and Glu72, respectively. This in turn forces a 90° rotation of the peptide linkage between residues 71 and 72, and a 180° flip of the peptide linkage between Glu72 and His73, resulting in a crankshaft-like rotation that substantially re-orients several side chains of the sensor loop (Fig. 4B). The side chain of Ile71 adapts a new rotamer conformation 120° from that in the ADP-bound state, with little repercussion. Glu72, however, is transported angstroms away from its position in the ADP-state (2.3 Å and 4.1 Å shifts of its {alpha}- and {delta}-carbons), and similarly for His73 (1.1 Å and 1.9 Å for the {alpha}-carbon and distal end of the imidazole ring.).


Figure 3
View larger version (33K):
[in this window]
[in a new window]
  		FIGURE 3.
A, crystal structures of AP-actin in the ADP-bound (blue) and ATP-bound (orange) states show structural differences in the active site and sensor loop, but essentially no significant differences elsewhere. Likewise, comparison of AP-actin with TMR-modified actin in either the (B) AMPPNP-(green) or (C) ADP-states (red) show no significant overall structural differences, other than the absence of an ordered helix in the AP-actin·ADP state. In all three panels, only {alpha}-carbons of subdomains 3 and 4 were used in the superpositions, to highlight any potential rotations in subdomains 1 and 2 relative to 3 and 4.

 
These structural perturbations of the sensor loop impact multiple residues in its immediate vicinity. In the ATP-bound state, for example, Glu72 forms a new bond with Thr77, which applies a slight torque to the threonine peptide backbone, inciting a 1.1-Å shift in the carboxamide side chain of Asn78 on the protein surface, but propagating no further than an 0.4-Å shift in Trp79, also on the protein surface. The bonds made by the side chain of Arg183 (in subdomain 4) with Glu72 and the backbone carbonyl between residues 72 and 73 in the ADP-bound state are abolished when ATP binds, leaving the arginine to preferentially occupy an alternate conformation (Fig. 6). His73 forms a tighter hydrogen bond with the backbone carbonyl between residues 158 and 159 in domain 3, from 3.1 Å in the ADP-state to 2.9 Å in the ATP-state. The methyl group expected on His73 is not clearly seen in the AP-actin crystal structures, most likely because it is not present at full occupancy, as determined by mass spectroscopy (see discussion in Ref. 24.) To better stack its carboxylate against His73, Asp179 moves 0.6 Å closer to it, which subsequently induces an 0.7 Å reorientation of the guanidinium of Arg177 to improve salt-bridge formation with Asp179.

Notably, all of these changes affect only the surface of the protein, and do not propagate further into the domains. Fig. 4C shows that while the changes propagate to a small extent toward the subdomain 2 region, they dissipate by Arg37. All of these shifts, albeit small, are supported by the isomorphous difference Fourier maps. Most of the changes are obligatory, in that strong Van der Waal repulsion arising from the presence of the {gamma}-phosphate of ATP requires these responses. In both the ATP- and ADP-bound states the active site calcium ion is coordinated in a pentagonal-bipyramidal manner, with one axial ligand provided by the beta-phosphate of the nucleotide. In the ADP-bound state, the other six ligands are water molecules, one of which is displaced by an oxygen of the {gamma}-phosphate in the ATP-bound state. Aspartates 11 and 154 ligate four water molecules of the calcium hydration shell.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Here we present the first crystal structures of a monomeric actin, in the ADP and ATP states, which are neither complexed with other proteins or toxins nor chemically cross-linked to prevent polymerization. Several biochemical assays (ATP hydrolysis, nucleotide exchange, subtilisin digestion, cofilin binding) were used to establish that the effect of the two surface mutations on actin structure were restricted to their effects on polymerization, the intended goal of the mutations. The crystal structures showed that significant and obligatory conformational changes are caused by the presence of the {gamma}-phosphate in the ATP state. These changes present a substantially different interaction potential on the surface of the actin monomer (Fig. 6).

Conformation of the D-loop—Our data do not support the hypothesis that the D-loop in subdomain 2 forms an {alpha}-helix in the ADP but not the ATP state (2). The isomorphous difference Fourier map (Fig. 4C) clearly reveals that there are no changes in electron density in the D-loop region, and thus there is no formation of an ordered {alpha}-helix in this region in the ADP state of AP-actin. Why does our ADP-bound AP-actin structure not show formation of an ordered short {alpha}-helix as is seen in the D-loop of ADP-bound TMR-actin (Fig. 3C)? A significant difference between the two studies is the nature of the crystal packing contacts. Our structure shows very little contact with the neighboring molecule in the region of subdomain 2 (Fig. 7A) in the crystal. In contrast, as pointed out by Sablin et al. (4), the ADP crystal of TMR-actin shows remarkably fortuitous intermolecular crystal contacts that could stabilize formation of a short {alpha}-helical segment (Fig. 7B). Because of the abundance of glycine residues, the sequence of residues 40-48 comprising the putative helical region, HQGVMVGMG, would not be expected to be particularly conducive to {alpha}-helix formation. The fact that a helix was not seen in the AMPPNP structure of TMR-actin, which had the same crystal packing as the ADP structure, was considered strong validation that this nucleotide-dependent conformational change was real.


Figure 4
View larger version (32K):
[in this window]
[in a new window]
  		FIGURE 4.
A, isomorphous difference Fourier map between actin·ATP and actin·ADP (Fobs,actin. ATP - Fobs,actin.ADP; positive difference density contoured at +4 sigma in green, -4 sigma in red) covering the entire protein shows changes localized to the nucleotide binding region and sensor loop. Positive density (green contours) indicates regions where the density in the ATP state is greater than in the ADP state, and vice versa for negative density (red contours). B, close-up of the nucleotide binding region and sensor loop (lower circle in A) unambiguously confirms the shifts in Ser14, Glu72, and His73, and the 90° and 180° rotations of the peptide linkages between residues 71-72 and 72-73, respectively. C, close-up of the region of the upper circle in A, at a lower contour level of ±2.5 sigma, shows that any propagated structural rearrangements vanish at the distal end of the beta-sheet before reaching the D-loop region.

 


Figure 5
View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 5.
Hydrogen-bonding and salt-bridge interactions of AP-actin with (A) ATP and (B) ADP. Residues of subdomains 1 and 2 are colored red, those of subdomains 3 and 4 across the nucleotide cleft are green. Bonding distances (in Å) are only listed in B if they differ from those in A.

 


Figure 6
View larger version (54K):
[in this window]
[in a new window]
  		FIGURE 6.
Nucleotide-dependent conformational changes in Glu72 and His73 in the sensor loop, and Arg183 in subdomain 4 present a substantially different surface on the exterior of the actin monomer. The nucleotide shown in green is for reference only and is not on or part of the surface.

 


Figure 7
View larger version (19K):
[in this window]
[in a new window]
  		FIGURE 7.
Crystal packing interactions between the actin monomer (black {alpha}-carbon trace) and its neighbors (green) in the region of domain 2 in (A) AP-actin and (B) TMR-actin crystals. While there is plenty of open space for an {alpha}-helix to form unhindered in the AP-actin crystals, no helix is seen. In contrast, the {alpha}-helix, seen only in the ADP-bound state of TMR-modified actin (red arrowhead), is impacted quite extensively by adjacent monomers in the crystal.

 
Re-examination of the TMR-modified actin structures, however, suggests that the nucleotide-dependent formation of the D-loop helix may result from signal propagation through crystal packing interactions. Fig. 8A indicates backbone atoms (in red) of the TMR-modified actin structures that differ by more than 0.4 Å between the ADP and ATP states, after superposition of the two structures. As expected, differences are seen in subdomain 2, but surprisingly significant and contiguous differences are also seen in the beta-ribbon extending from residues 237 to 252 of domain 4, far removed from the active site and subdomain 2. In the crystalline environment of TMR-actin, however (Fig. 8B), the nucleotide-dependent changes in the sensor loop result in a 3.1-Å hydrogen bond being formed between Glu72 and Gln246 of an adjacent monomer within the crystal in the ADP state. This interaction tugs on the cantileverlike beta-ribbon of the adjacent monomer. As a result of this shift, at the end of this ribbon Leu236 reorients in the ADP state to better pack against Gly251. This reorientation of Leu236 further amplifies conformational rear-rangement in the loop between residues 231 and 236. In the ATP state, Ala231 of this loop protrudes into the space occupied by the D-loop helix of yet a third adjacent monomer within the crystal, and thus sterically precludes formation of the D-loop {alpha}-helix. Ala231 is retracted from this space in the ADP state, removing the steric block and allowing the D-loop {alpha}-helix (green, visible only in the ADP state) to form. This crystal packing-mediated pathway appears to be equally plausible to any proposed intramolecular pathway for formation of the ordered D-loop helix, but is unlikely to be physiologically relevant.

One might question whether Ala231 is retracted as a result of the cantilever motion propagated from the tug imparted by the bond formed between Glu72 and Gln246 in the ADP state, or whether formation of the D-loop helix simply pushes Ala231 out of the way. Indeed, the beta-ribbon of the cantilever superimposes rather well in the two states (Fig. 8B), the differences being approximately at the level of error in the models coordinates. However, the isomorphous difference Fourier map made between the observed diffraction amplitudes of the TMR-actin-ADP and -ATP crystals reveals the characteristic peak-hole pairs that clearly indicate a concerted directional shift of the cantilever. In fact, it was through detection of this pattern in the isomorphous difference map that led us to propose the cantilever mechanism. This does not prove but strongly implicates the cantilever as the agent that couples the events occurring at its ends. Because both the AP-actin and TMR-actin structures suggest that the re-positioning of Glu72 in response to a change in nucleotide state is primary (causative), it follows that the protrusion or retraction of Ala231 is the subsequent response.

Disposition of the Nucleotide Binding Cleft—The second hypothesis that has been proposed to account for the anticipated large changes between ADP and ATP actin is an opening of the nucleotide cleft dividing the two major domains (1-2 and 3-4) upon ATP hydrolysis. This strategy has been employed by a host of other nucleotide-hydrolyzing proteins, including motor proteins, G-proteins, and hexokinase (reviewed in Ref. 4). Crystallographically, the only actin structure to show an open cleft is an actin-profilin complex (25). An open cleft is defined as an increased separation of {approx}3 Å between two beta-hairpins that are located on either side of the nucleotide (residues 11-16 in subdomain 1 and residues 154-161 in subdomain 3) at the base of the nucleotide cleft. Recent molecular dynamic simulations show that the open nucleotide cleft of actin-profilin reverts to its preferred closed cleft conformation once the profilin is removed from the simulation (26). The authors conclude from this molecular dynamics study that there is no thermodynamically stable open cleft structure of monomeric actin.


Figure 8
View larger version (24K):
[in this window]
[in a new window]
  		FIGURE 8.
Potential signal propagation through crystal packing interactions in TMR-actin. A, backbone coordinates of TMR-actin that differ by more than 0.4 Å between the ADP- and AMPPNP-bound states are indicated in red. The region comprising residues 231-252 undergoes a shift between the two nucleotide states despite its distance from the active site region. B, slice through the TMR-actin crystals showing three adjacent monomers reveals a potential mechanism by which nucleotide-dependent changes may be relayed through the crystal to the D-loop region of a symmetry-related TMR-actin monomer. The beta-ribbon (residues 237-252) of the center monomer (ADP state in blue, ATP state in red) senses the ADP state through a hydrogen bond that forms between its Gln246 and the Glu72 of an adjacent monomer in the crystal. The beta-ribbon cantilever relays the shift to Leu236, which reorients under the strain to improve its hydrophobic packing and thus amplifies the signal. This results in a conformational rearrangement extending to Ala231, which is retracted in the ADP state to remove the steric block that prevents formation of an {alpha}-helix in the D-loop (green) in the ATP-like state.

 
In the structures of both AP-actin and TMR-actin, the cleft is closed whether ADP or ATP is present. The position of the adenine ring, ribose, and {alpha}- and beta-phosphates is the same in both nucleotide states, and is stabilized by residues on either side of the cleft that divides the two major domains. Thus there is no structural reason why hydrolysis of ATP should loosen the connection between the two halves of the molecule and cause an opening of the cleft, in the absence of other protein-protein contacts. It has been suggested that the ADP form of the TMR-actin structure was prohibited from adopting an open cleft by the presence of the rhodamine in the cleft between subdomain 1 and 3. Our AP-actin structure with bound ADP also showed a closed cleft, yet it is not subject to this criticism, because the two point mutations are located on the surface of subdomain 4.


Figure 9
View larger version (38K):
[in this window]
[in a new window]
  		FIGURE 9.
Comparison of nucleotide-dependent domain motions in AP-actin versus TMR-actin. Vectors indicating the direction of atomic shifts in going from the ADP state (red end of vectors) to the ATP state (green end) are shown for both AP-actin (middle panels) and TMR-actin (bottom panels). This figure was made by superimposing the C{alpha} coordinates of domains 3 and 4 of each protein in the ADP and ATP state, and generating the vector between each C{alpha} atom from the ADP to ATP state. Because nearly all of the resulting vectors would be too small to see clearly, their magnitudes are shown exaggerated by 10-fold to make visible the subtle conformational reorientations. The red ends of the vectors are anchored on the C{alpha} atoms of the ADP state. Shown from two viewpoints, the small shifts in going from the ADP-state to the ATP-state in AP-actin are opposite to those of the same transition in TMRactin, suggesting that these slight global motions are not obligatory coupled to the change in nucleotide state.

 
What structural evidence supports an open cleft structure for ADP-actin? Electron microscopic reconstructions of negatively stained yeast F-actin filaments in the ADP state were best fitted with the open-cleft actin crystal structure, while the closedcleft structure provided a better fit in the presence of the ATP analog ADP·beryllium fluoride (5). It remains to be determined if similar results hold true for vertebrate actin, because yeast actin shows only 87% sequence identity with striated muscle actins. A second line of evidence consistent with the idea that ADP-actin might form an open cleft comes from the structure of the actin-related protein 2/3 (Arp2/3) complex, which consists of seven proteins, two of which are homologous to actin. The Arp2/3 complex was originally crystallized in the apo state in which Arp3 had an open cleft (27). When ATP was soaked into these crystals, the nucleotide binding cleft of Arp3 closed, while addition of ADP caused only modest changes (28). Notably, the ADP in Arp3 bound in a unique conformation that is incompatible with the closed conformation seen with ATP.

Domain Motions—Another feature of actin that could allow ADP- and ATP-bound forms to be discriminated from one another is the relative orientation of the two large domains (1 and 2 versus 3 and 4) with respect to one another. To get a sense of concerted changes in the molecule going from the ADP to the ATP state, subdomains 3 and 4 of AP-actin were superimposed to show the motions that result in subdomains 1 and 2. Fig. 9 illustrates this motion with vectors that indicate the direction of atomic shifts starting from the ADP state of AP-actin (red end of shift vectors) and pointing toward the ATP state (green end). Note that the magnitude of the changes are exaggerated 10-fold to make visible the otherwise subtle displacements. The most striking feature of this analysis is that the motions are all in one direction, giving rise to a counter-clockwise rotation of subdomains 1 and 2 in the standard view of the monomer. In contrast, a similar analysis with TMR-actin shows domain motion in the opposite sense as AP-actin. Quantitatively, the vector correlation in the nucleotide-dependent shift of domains 1 and 2 between AP- and TMR-actins is -0.09, indicating a slight overall anti-correlation of motions. Therefore there is no consistent evidence for global movement of the two halves of the molecule relative to one another that is obligatorily coupled to a change in nucleotide state.

How can the absence of significant domain motions or conformational changes in the D-loop between the ADP- and ATP-bound states be reconciled with the solution experiments that suggest otherwise? Given that there are very few crystal packing interactions involving the D-loop region of our AP-actin structures (Fig. 7A) that would prevent expression of any conformational changes there, the obligatory nucleotide-dependent conformational changes in the isolated actin monomer appear to be confined to the active site and sensor loop region. So how can the D-loop be more susceptible to subtilisin cleavage in the ATP state than in the ADP state if there are no structural differences seen in that region? One possible explanation is that the ATP-bound monomer forms transient dimers even at low concentrations in G-buffer, which is consistent with ATP-monomers having higher affinity for each other than ADP-monomers. We propose that in this transient dimeric state the D-loop region of domain 2 undergoes a conformational change, and that in this state Met47 and Gly48 of the D-loop become better disposed for subtilisin cleavage.

It is clear that the D-loop region is structurally plastic, as evidenced by the variety of conformations it has been seen to adopt. These include a disordered loop as seen here, an {alpha}-helix in the TMR-actin ADP structure (2), and a beta-sheet in the actin-DNase I structure (29). Dominguez et al. (30) propose that a helical D-loop of one actin monomer binds in the hydrophobic cleft of the adjacent actin monomer in the F-actin helix. This suggestion would require some modification of the prevailing Holmes model of the filament (31). The hydrophobic pocket, located between subdomains 1 and 3, is the site at which helices from a number of other actin-binding proteins also bind (reviewed in Ref. 30), and is coincident with the binding site of several toxins that interfere with polymerization (32).

Alternatively, it is possible that even relatively weak crystal packing forces may prevent expression of some larger conformational change. It has been suggested that open and closed states of some proteins that undergo significant domain motions may differ only slightly in energy (33). If a low energy barrier separates the various conformations of actin, then one would expect to see a range of conformations expressed in the multitude of actin structures due to their diverse crystal packing arrangements. In fact, all actin structures look remarkably the same, despite the myriad ways this protein is arranged within the crystals, with two exceptions: the complex in which actin is bound to profilin shows an open cleft (25), and the ADP structure of TMR-actin (2), wherein the D-loop forms an {alpha}-helical segment. It is thus most likely that domains 1, 3, and 4 remain rigidly disposed relative to each other independent of nucleotide state, and that the conformation of domain 2 depends upon interactions with other proteins (usually the adjacent actin in the filament.) Since crystals of physiological dimers of actin remain elusive, so does the physiologically-relevant structure of the D-loop of domain 2.

Active Site—We have suggested that the full range of nucleotide induced changes in actin may not be realized until the monomer binds to another actin monomer or is incorporated into the filament. An example of this is illustrated by the several thousand-fold increase in rate of nucleotide hydrolysis in the filament compared with the free monomer (34). The mechanism for this enhancement is not known, but it has been speculated that actin-actin contacts may promote small rearrangements of active site residues that produce a more catalytically active conformation (20), similar to the mechanism proposed for activation of trimeric G proteins by its regulators (reviewed in Ref. 35). This allosteric view is favored because existing models of the actin filament do not allow for a direct contribution of a catalytic residue from a neighboring monomer.

Implications for the Actin Filament—A cross-linked longitudinal actin dimer prepared from tissue-purified actin was recently crystallized, with the goal of beginning to obtain a high resolution structure of actin in the filamentous state (36). In addition to considerable disorder in subdomain 2, the crystal also contained only one monomer per asymmetric unit, and thus the cross-linked region could not be unambiguously visualized. Whereas no cross-linking agents were used in either of the AP-actin crystals, we observed "longitudinal" interactions (crystal contacts) between subdomain 3 and an adjacent subdomain 4 that are nearly identical to those observed in the crosslinked dimer crystal intended to capture these interactions (36). It is not clear whether these weak interactions, seen with both ADP-and ATP-bound AP-actin, represent a step toward the formation of an actin filament.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
In the absence of chemical cross-links, bound proteins or toxins, crystals of constitutively monomeric AP-actin in the ADP- and ATP-bound states confirm that obligatory nucleotide-dependent changes in structure occur in the active site region and sensor loop on the protein's surface. No significant conformational changes propagate elsewhere in our monomeric crystal structures. In contrast to the TMR-modified actin structures, the D-loop does not become an ordered {alpha}-helix upon nucleotide hydrolysis. However, it is clear from the multitude of actin crystal structures that this region of the molecule has the potential to adopt a variety of conformations. The full expression of nucleotide-dependent conformational changes in actin may require stabilization via actin-actin contacts, because the structure of an actin monomer within a filament will likely differ from that of an unpolymerized monomer.


   FOOTNOTES
 
The atomic coordinates and structure factors (code 2HF3 and 2HF4) have been deposited in the Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

* This work was supported by National Institutes of Health Grant HL38113 (to K. M. T.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 To whom correspondence should be addressed: Dept. of Molecular Physiology and Biophysics, University of Vermont, Burlington, VT 05405. Tel.: 802-656-8750; Fax: 802-656-0747; E-mail: kathleen.trybus{at}uvm.edu.

2 The abbreviations used are: TMR-actin, tetramethylrhodamine-Cys374-labeled monomeric actin; AP-actin, actin-A204E/P243K; {epsilon}-ATP, etheno-ATP; DTT, dithiothreitol; RMSD, root mean square deviation; AMPPNP, adenosine 5'-(beta,{gamma}-imino)triphosphate. Back


   ACKNOWLEDGMENTS
 
We thank Roberto Dominguez for providing observed structure factors for the TMR-actin crystals in the ADP and AMPPNP states.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 

  1. Dominguez, R., and Graceffa, P. (2003) Biophys. J. 85, 2073-2074[Medline] [Order article via Infotrieve]
  2. Otterbein, L. R., Graceffa, P., and Dominguez, R. (2001) Science 293, 708-711[Abstract/Free Full Text]
  3. Graceffa, P., and Dominguez, R. (2003) J. Biol. Chem. 278, 34172-34180[Abstract/Free Full Text]
  4. Sablin, E. P., Dawson, J. F., VanLoock, M. S., Spudich, J. A., Egelman, E. H., and Fletterick, R. J. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 10945-10947[Free Full Text]
  5. Belmont, L. D., Orlova, A., Drubin, D. G., and Egelman, E. H. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 29-34[Abstract/Free Full Text]
  6. Page, R., Lindberg, U., and Schutt, C. E. (1998) J. Mol. Biol. 280, 463-474[CrossRef][Medline] [Order article via Infotrieve]
  7. Volkman, L., Storm, K., Aivazachvili, V., and Oppenheimer, D. (1996) Virology 225, 369-376[CrossRef][Medline] [Order article via Infotrieve]
  8. Joel, P. B., Fagnant, P. M., and Trybus, K. M. (2004) Biochemistry 43, 11554-11559[CrossRef][Medline] [Order article via Infotrieve]
  9. Pardee, J. D., and Spudich, J. A. (1982) Methods Enzymol. 85, 164-181
  10. Pollard, T. D. (1986) J. Cell Biol. 103, 2747-2754[Abstract/Free Full Text]
  11. Kudryashov, D. S., and Reisler, E. (2003) Biophys. J. 85, 2466-2475[Medline] [Order article via Infotrieve]
  12. Kissinger, C. R., Gehlhaar, D. K., and Fogel, D. B. (1999) Acta Crystallogr. D. Biol. Crystallogr. 55, 484-491[CrossRef][Medline] [Order article via Infotrieve]
  13. Hodel, A., Kim, S., and Brunger, A. (1992) Acta Crystallographica Section A 48, 851-858[CrossRef]
  14. Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) Acta Crystallogr. D. Biol. Crystallogr. 54, 905-921[CrossRef][Medline] [Order article via Infotrieve]
  15. Brunger, A. T. (1997) Methods Enzymol. 277, 366-396[CrossRef][Medline] [Order article via Infotrieve]
  16. Rould, M. A., and Carter, C. W. (2003) Methods Enzymol. 374, 145-163[Medline] [Order article via Infotrieve]
  17. Rould, M. A. (2006) Methods Mol. Biol. 364, in press
  18. Kelleher, J. F., Atkinson, S. J., and Pollard, T. D. (1995) J. Cell Biol. 131, 385-397[Abstract/Free Full Text]
  19. Carlier, M. F. (1990) Adv. Biophys. 26, 51-73[CrossRef][Medline] [Order article via Infotrieve]
  20. Vorobiev, S., Strokopytov, B., Drubin, D. G., Frieden, C., Ono, S., Condeelis, J., Rubenstein, P. A., and Almo, S. C. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 5760-5765[Abstract/Free Full Text]
  21. Kudryashov, D. S., Phillips, M., and Reisler, E. (2004) Biophys. J. 87, 1136-1145[CrossRef][Medline] [Order article via Infotrieve]
  22. Strzelecka-Golaszewska, H., Moraczewska, J., Khaitlina, S. Y., and Mossakowska, M. (1993) Eur. J. Biochem. 211, 731-742[Medline] [Order article via Infotrieve]
  23. Carlier, M. F., Laurent, V., Santolini, J., Melki, R., Didry, D., Xia, G. X., Hong, Y., Chua, N. H., and Pantaloni, D. (1997) J. Cell Biol. 136, 1307-1322[Abstract/Free Full Text]
  24. Bookwalter, C. S., and Trybus, K. M. (2006) J. Biol. Chem. 281, 16777-16784[Abstract/Free Full Text]
  25. Chik, J. K., Lindberg, U., and Schutt, C. E. (1996) J. Mol. Biol. 263, 607-623[CrossRef][Medline] [Order article via Infotrieve]
  26. Minehardt, T. J., Kollman, P. A., Cooke, R., and Pate, E. (2006) Biophys. J. 90, 2445-2449[CrossRef][Medline] [Order article via Infotrieve]
  27. Robinson, R. C., Turbedsky, K., Kaiser, D. A., Marchand, J. B., Higgs, H. N., Choe, S., and Pollard, T. D. (2001) Science 294, 1679-1684[Abstract/Free Full Text]
  28. Nolen, B. J., Littlefield, R. S., and Pollard, T. D. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 15627-15632[Abstract/Free Full Text]
  29. Kabsch, W., Mannherz, H. G., Suck, D., Pai, E. F., and Holmes, K. C. (1990) Nature 347, 37-44[CrossRef][Medline] [Order article via Infotrieve]
  30. Dominguez, R. (2004) Trends Biochem. Sci. 29, 572-578[CrossRef][Medline] [Order article via Infotrieve]
  31. Holmes, K. C., Popp, D., Gebhard, W., and Kabsch, W. (1990) Nature 347, 44-49[CrossRef][Medline] [Order article via Infotrieve]
  32. Klenchin, V. A., Allingham, J. S., King, R., Tanaka, J., Marriott, G., and Rayment, I. (2003) Nat. Struct. Biol. 10, 1058-1063[CrossRef][Medline] [Order article via Infotrieve]
  33. Simanshu, D. K., Savithri, H. S., and Murthy, M. R. (2005) J. Mol. Biol. 352, 876-892[CrossRef][Medline] [Order article via Infotrieve]
  34. Pollard, T. D., and Weeds, A. G. (1984) FEBS Lett. 170, 94-98[CrossRef][Medline] [Order article via Infotrieve]
  35. Ross, E. M., and Wilkie, T. M. (2000) Annu. Rev. Biochem. 69, 795-827[CrossRef][Medline] [Order article via Infotrieve]
  36. Kudryashov, D. S., Sawaya, M. R., Adisetiyo, H., Norcross, T., Hegyi, G., Reisler, E., and Yeates, T. O. (2005) Proc. Natl. Acad. Sci. U. S. A. 102, 13105-13110[Abstract/Free Full Text]
  37. Luzzati, V. (1952) Acta Crystallogr. 5, 802-810[CrossRef]
  38. Laskowski, R., MacArthur, M., Moss, D., and Thornton, J. (1993) J. Appl. Crystallogr. 26, 283-291[CrossRef]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
T. D. Pollard and J. Berro
Mathematical Models and Simulations of Cellular Processes Based on Actin Filaments
J. Biol. Chem., February 27, 2009; 284(9): 5433 - 5437.
[Abstract] [Full Text] [PDF]

Home page
 JCBHome page
V. O. Paavilainen, E. Oksanen, A. Goldman, and P. Lappalainen
Structure of the actin-depolymerizing factor homology domain in complex with actin
J. Cell Biol., October 23, 2008; 182(1): 51 - 59.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
K. Baek, X. Liu, F. Ferron, S. Shu, E. D. Korn, and R. Dominguez
Modulation of actin structure and function by phosphorylation of Tyr-53 and profilin binding
PNAS, August 19, 2008; 105(33): 11748 - 11753.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
P. Kapoor, A. A. Sahasrabuddhe, A. Kumar, K. Mitra, M. I. Siddiqi, and C. M. Gupta
An Unconventional Form of Actin in Protozoan Hemoflagellate, Leishmania
J. Biol. Chem., August 15, 2008; 283(33): 22760 - 22773.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
M. Iwasa, K. Maeda, A. Narita, Y. Maeda, and T. Oda
Dual Roles of Gln137 of Actin Revealed by Recombinant Human Cardiac Muscle {alpha}-Actin Mutants
J. Biol. Chem., July 25, 2008; 283(30): 21045 - 21053.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
B. M. Miller and K. M. Trybus
Functional Effects of Nemaline Myopathy Mutations on Human Skeletal {alpha}-Actin
J. Biol. Chem., July 11, 2008; 283(28): 19379 - 19388.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
E. Reisler and E. H. Egelman
Actin Structure and Function: What We Still Do Not Understand
J. Biol. Chem., December 14, 2007; 282(50): 36133 - 36137.
[Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
I. Fujiwara, D. Vavylonis, and T. D. Pollard
Polymerization kinetics of ADP- and ADP-Pi-actin determined by fluorescence microscopy
PNAS, May 22, 2007; 104(21): 8827 - 8832.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
281/42/31909    most recent
M601973200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rould, M. A.
Right arrow Articles by Trybus, K. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rould, M. A.
Right arrow Articles by Trybus, K. M.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2006 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement