Portal de Periódicos da CAPES

The Light Chain Binding Domain of Expressed Smooth Muscle Heavy Meromyosin Acts as a Mechanical Lever

2000; Elsevier BV; Volume: 275; Issue: 47 Linguagem: Inglês

10.1074/jbc.m006438200

1083-351X

David M. Warshaw, William H. Guilford, Yelena Freyzon, Elena B. Krementsova, Kimberly A. Palmiter, Mathew J. Tyska, Josh E. Baker, Kathleen M. Trybus,

Cellular Mechanics and Interactions

Structural data led to the proposal that the molecular motor myosin moves actin by a swinging of the light chain binding domain, or "neck." To test the hypothesis that the neck functions as a mechanical lever, smooth muscle heavy meromyosin (HMM) mutants were expressed with shorter or longer necks by either deleting or adding light chain binding sites. The mutant HMMs were characterized kinetically and mechanically, with emphasis on measurements of unitary displacements and forces in the laser trap assay. Two shorter necked constructs had smaller unitary step sizes and moved actin more slowly than WT HMM in the motility assay. A longer necked construct that contained an additional essential light chain binding site exhibited a 1.4-fold increase in the unitary step size compared with its control. Kinetic changes were also observed with several of the constructs. The mutant lacking a neck produced force at a somewhat reduced level, while the force exerted by the giraffe construct was higher than control. The single molecule displacement and force data support the hypothesis that the neck functions as a rigid lever, with the fulcrum for movement and force located at a point within the motor domain. Structural data led to the proposal that the molecular motor myosin moves actin by a swinging of the light chain binding domain, or "neck." To test the hypothesis that the neck functions as a mechanical lever, smooth muscle heavy meromyosin (HMM) mutants were expressed with shorter or longer necks by either deleting or adding light chain binding sites. The mutant HMMs were characterized kinetically and mechanically, with emphasis on measurements of unitary displacements and forces in the laser trap assay. Two shorter necked constructs had smaller unitary step sizes and moved actin more slowly than WT HMM in the motility assay. A longer necked construct that contained an additional essential light chain binding site exhibited a 1.4-fold increase in the unitary step size compared with its control. Kinetic changes were also observed with several of the constructs. The mutant lacking a neck produced force at a somewhat reduced level, while the force exerted by the giraffe construct was higher than control. The single molecule displacement and force data support the hypothesis that the neck functions as a rigid lever, with the fulcrum for movement and force located at a point within the motor domain. essential light chain regulatory light chain heavy meromyosin wild type β-cardiac actin binding loop mean-variance Muscle contracts as a result of the cyclic interaction of the molecular motor myosin with actin, powered by the hydrolysis of MgATP. A simple mechanistic model by which myosin could move actin was proposed based on the crystal structure of skeletal myosin subfragment 1 (1Rayment I. Rypniewski W.R. Schmidt-Base K. Smith R. Tomchick D.R. Benning M.M. Winkelmann D.A. Wesenberg G. Holden H.M. Science. 1993; 261: 50-58Crossref PubMed Scopus (1877) Google Scholar, 2Rayment I. Holden H.M. Whittaker M. Yohn C.B. Lorenz M. Holmes K.C. Milligan R.A. Science. 1993; 261: 58-65Crossref PubMed Scopus (1456) Google Scholar, 3Fisher A.J. Smith C.A. Thoden J. Smith R. Sutoh K. Holden H.M. Rayment I. Biophys. J. 1995; 68: 19S-26SPubMed Google Scholar). The key feature was an 8.5-nm single α-helix, stabilized by the essential and regulatory light chains (ELC and RLC),1 which formed an elongated neck region that emerged from the globular motor domain. It was suggested that a substantial portion of the myosin motor domain maintains a fixed orientation when attached to actin, while the neck region pivots about a fulcrum within the motor domain, thus generating a power stroke. Additional evidence in support of a lever arm rotation was obtained from the crystal structure of a motor domain-essential light chain complex with a transition state analog at the active site, which showed the lever arm in a second position that may represent myosin in the prepowerstroke state (4Dominguez R. Freyzon Y. Trybus K.M. Cohen C. Cell. 1998; 94: 559-571Abstract Full Text Full Text PDF PubMed Scopus (591) Google Scholar). The skeletal subfragment 1 structure is likely to resemble the structure adopted at the end of the powerstroke. A comparison of the two conformations shows that smaller changes that originate at the active site are amplified into much larger movements of the lever arm. This motion could accommodate a powerstroke on the order of 10 nm, in the range of the 5–15 nm of displacement measured using single molecule techniques (5Finer J.T. Simmons R.M. Spudich J.A. Nature. 1994; 368: 113-119Crossref PubMed Scopus (1583) Google Scholar, 6Molloy J.E. Burns J.E. Kendrick-Jones J. Tregear R.T. White D.C.S. Nature. 1995; 378: 209-212Crossref PubMed Scopus (527) Google Scholar, 7Guilford W.H. Dupuis D.E. Kennedy G. Wu J. Patlak J.B. Warshaw D.M. Biophys. J. 1997; 72: 1006-1021Abstract Full Text PDF PubMed Scopus (223) Google Scholar, 8Kitamura K. Tokunaga M. Iwane A.H. Yanagida T. Nature. 1999; 397: 129-134Crossref PubMed Scopus (424) Google Scholar). The simplest mechanical model for the neck region predicts that myosin with a shorter neck (i.e. shorter lever arm) should generate smaller unitary displacements and move actin more slowly, whereas a longer neck should lead to larger displacements and more rapid actin movement (reviewed in Ref. 9Block S. Cell. 1996; 87: 151-157Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar). Several studies in which myosins of various neck lengths were produced either by removing light chains (10Lowey S. Waller G.S. Trybus K.M. Nature. 1993; 365: 454-456Crossref PubMed Scopus (268) Google Scholar, 11Trybus K.M. J. Biol. Chem. 1994; 269: 20819-20822Abstract Full Text PDF PubMed Google Scholar) or by genetically adding or deleting light chain binding sites (12Uyeda T.Q.P. Spudich J.A. Science. 1993; 262: 1867-1870Crossref PubMed Scopus (103) Google Scholar, 13Uyeda T.Q. Abramson P.D. Spudich J.A. Proc. Natl. Acad. Sci. 1996; 93: 4459-4464Crossref PubMed Scopus (392) Google Scholar) showed that constructs with necks shorter than wild type moved actin more slowly, while a construct with a longer neck moved actin more quickly. A chimera in which two α-actinin repeats were fused to the Dictyostelium motor domain showed a higher average velocity than a similar construct with only one α-actinin repeat, suggesting that nonnative structures can mimic some aspects of the native neck (14Anson M Geeves M.A. Kurzawa S.E. Manstein D.J. EMBO J. 1996; 15: 6069-6074Crossref PubMed Scopus (136) Google Scholar). Although these studies are consistent with a simple lever arm model, they all relied on the assumption that no kinetic changes resulted from these biochemical or genetic perturbations. Since velocity in the motility assay,v max, is dependent upon both step size (d) and the time spent attached to actin following the powerstroke (t on) (i.e. v max ≅d/t on) (15Huxley H.E. J. Biol. Chem. 1990; 265: 8347-8350Abstract Full Text PDF PubMed Google Scholar), changes in either parameter could equally well account for the observed differences in motility. Here we test the lever arm hypothesis at the single molecule level by measuring the displacement (d) and force (F) of a series of smooth muscle heavy meromyosin (HMM) mutants in which light chain binding sites were either added to or deleted from the neck. The laser trap data support the hypothesis that the neck acts as a lever and are consistent with structural data that suggest that the fulcrum for movement and force is located near the SH1 helix (4Dominguez R. Freyzon Y. Trybus K.M. Cohen C. Cell. 1998; 94: 559-571Abstract Full Text Full Text PDF PubMed Scopus (591) Google Scholar). Wild type (WT) smooth muscle HMM (amino acids 1–1175) and neck length mutants of this heavy chain backbone were co-expressed with the regulatory and essential light chains using the baculovirus insect cell expression system (11Trybus K.M. J. Biol. Chem. 1994; 269: 20819-20822Abstract Full Text PDF PubMed Google Scholar). The HMM backbone was chosen as the basis for the different neck-length constructs so that monoclonal anti-rod antibody S2.2 could be used as a common means of attaching the molecules to the nitrocellulose substratum for both motility and laser trap studies. Proteins were purified by binding to actin and release with MgATP as described previously (11Trybus K.M. J. Biol. Chem. 1994; 269: 20819-20822Abstract Full Text PDF PubMed Google Scholar). Two constructs with neck lengths shorter than WT HMM were cloned (see Fig. 1). An HMM with no light chain binding region ("neckless") had heavy chain residues 791–848 deleted. This resulted in a dimeric construct with the motor domain attached to the rod. The sequence of the region joining the motor domain to the rod in the neckless construct was ERDLGPLLQV. An HMM mutant lacking an RLC binding site ("-R site") had amino acids 820–848 deleted. The sequence of the joining region in this construct, which contains the motor domain and ELC binding site attached to the rod, was QQQLLGPLLQV.Figure 1Schematic diagram of the constructs used in this study.View Large Image Figure ViewerDownload Hi-res image Download (PPT) A long necked mutant ("giraffe") in which a second ELC binding site was added between the native ELC and RLC binding sites was also cloned (Fig. 1). This mutant retains native contacts between the motor domain and the ELC and between the ELC and the RLC but introduces a foreign ELC-ELC interaction. The following sequence was introduced between Leu819 and Thr820: LGITDVIIAFQAQCRGYLARKAFAKRQQQL. This sequence is LG plus amino acids 792–819. To test if the orientation of the two ELCs with respect to each other influences the properties of the giraffe construct, a second variation was constructed ("giraffe's twisted sister"). This revised construct had one less amino acid in the added ELC site than the previous construct; thus, the two ELCs would be rotated by approximately 100°. This was accomplished by inserting the sequence of amino acids 791–819 (i.e. KITDVIIAFQAQCRGYLARKAFAKRQQQL) between Leu819 and Thr820. A phosphorylation-independent variant of giraffe HMM was also engineered. Based on earlier studies, substitution of the 50/20-kDa β-cardiac actin binding loop (CABL) for the native sequence (residues 626–653) activated unphosphorylated smooth muscle HMM (16Rovner A.S. Freyzon Y. Trybus K.M. J. Biol. Chem. 1995; 270: 30260-30263Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). Thus, the double mutants, CABL-giraffe and CABL-giraffe's twisted sister, were also constructed. The movement of these constructs was more consistent than the original giraffe construct, and thus the single-molecule studies could be more readily performed. Unphosphorylated CABL-HMM with a native neck was used as the control for the unphosphorylated long necked CABL-giraffe. Actin-activated ATPase assays were performed in 10 mm imidazole, pH 7, 8 mm KCl, 1 mmMgCl2, 1 mm EGTA, 1 mmdithiothreitol, 1 mm NaN3 at 37 °C. Inorganic phosphate was determined colorimetrically at six time points per actin concentration, using SDS to stop the reaction (17White H.D. Methods Enzymol. 1982; 85: 698-708Crossref PubMed Scopus (121) Google Scholar). The concentration of active heads was determined by NH4+-ATPase activity relative to a myosin standard (25 mm Tris, pH 7.5, at 37 °C, 0.4m NH4Cl, 2 mm EDTA, 0.2m sucrose, 1 mm dithiothreitol, 1 mg/ml bovine serum albumin). Rotary-shadowed platinum images were obtained in 0.5 m ammonium acetate, 66% glycerol and observed with a Philips EM301 electron microscope operated at 60 kV (18Trybus K.M. Lowey S. J. Biol. Chem. 1984; 259: 8564-8571Abstract Full Text PDF PubMed Google Scholar). Actin was decorated with the expressed constructs, negatively stained, and observed over holes on carbon-coated grids (19Craig R. Szent-Gyorgyi A.G. Beese L. Flicker P. Vibert P. Cohen C. J. Mol. Biol. 1980; 140: 35-55Crossref PubMed Scopus (126) Google Scholar). The motility assay was performed at 30 °C in 25 mm imidazole, pH 7.5, 25 mm KCl, 4 mm MgCl2, 1 mm EGTA, 0.5% methylcellulose, as described by Trybus and Chatman (20Trybus K.M. Chatman T.A. J. Biol. Chem. 1993; 268: 4412-4419Abstract Full Text PDF PubMed Google Scholar). CABL-giraffe was also assayed at a higher ionic strength in 25 mmimidazole, pH 7.5, 60 mm KCl, 4 mmMgCl2, 1 mm EGTA, 0.7% methylcellulose. Monoclonal anti-rod antibody S2.2 was used to ensure a uniform site of attachment to the nitrocellulose surface (21Trybus K.M. Henry L. J. Cell Biol. 1989; 109: 2879-2886Crossref PubMed Scopus (34) Google Scholar). Fifteen or more filaments were used to calculate the mean and S.D. of the velocity of movement. The average velocity for a given filament was calculated by dividing the distance traveled by the filament during a 1-s period between video snapshots for 7–15 consecutive snapshots as described previously (22Work S.S. Warshaw D.M. Anal. Biochem. 1992; 202: 275-285Crossref PubMed Scopus (55) Google Scholar). To control for sporadic motility potentially compromising the velocity estimate, velocities for a paired preparation of CABL-HMM and CABL-giraffe were determined without averaging over numerous short intervals (1 s) and then plotted in a histogram (see Fig. 4 C). Technical details of our laser trap techniques have been published elsewhere (7Guilford W.H. Dupuis D.E. Kennedy G. Wu J. Patlak J.B. Warshaw D.M. Biophys. J. 1997; 72: 1006-1021Abstract Full Text PDF PubMed Scopus (223) Google Scholar, 23Dupuis D.E. Guilford W.H. Wu J. Warshaw D.M. J. Muscle Res. Cell Motil. 1996; 18: 17-30Crossref Scopus (139) Google Scholar, 24Lauzon A.-M. Tyska M.J. Rovner A.S. Freyzon Y. Warshaw D.M. Trybus K.M. J. Muscle Res. Cell Motil. 1998; 19: 825-837Crossref PubMed Scopus (119) Google Scholar). Briefly, a pair of 1-μm polystyrene microspheres coated withN-ethylmaleimide-modified myosin were held in solution using two independent laser traps. A fluorescently labeled (tetramethylrhodamine isothiocyanate-phalloidin) actin filament was strung between the two microspheres and pulled taut (>4 piconewtons). The actin filament is then lowered onto the surface of a 2.0-μm diameter silica bead adhered to a coverslip. The silica bead acts as a platform on which a sparse coating of HMM is applied, so that on average only one HMM molecule interacts with the actin filament at any time. HMM was adhered to the nitrocellulose-coated surface through an anti-subfragment 2 antibody applied at 100 μg/ml (21Trybus K.M. Henry L. J. Cell Biol. 1989; 109: 2879-2886Crossref PubMed Scopus (34) Google Scholar) before surface blocking with bovine serum albumin (500 μg/ml). HMM was then applied in concentrations varying from 2.5–10 μg/ml and allowed to bind to the antibody for 2 min. Thus, HMM was bound to the surface with random orientation but at a defined point on the molecule. All experiments were performed at low ionic strength (25 mmKCl), limiting ATP concentration (10 μm), and 25 °C, to prolong the unitary event durations. To measure the displacements imparted to the actin filament by a single HMM molecule, the bright field image of one of the microsphere handles is projected onto a quadrant photodiode detector, providingx and y bead position data. Myosin's unitary displacements were recorded under "unloaded" conditions (0.02–0.04 piconewtons/nm/trap). To measure force, a nearly isometric condition was obtained under feedback control by increasing the effective trap stiffness approximately 100-fold. The feedback loop contained an acousto-optic modulator, which was used to deflect the laser trap and thus counterbalance the myosin-generated forces (7Guilford W.H. Dupuis D.E. Kennedy G. Wu J. Patlak J.B. Warshaw D.M. Biophys. J. 1997; 72: 1006-1021Abstract Full Text PDF PubMed Scopus (223) Google Scholar). The acousto-optic modulator control signal was calibrated and served as the force signal. Force measurements in this assay may not represent the true maximum unitary force due to compliance in the actin filament-bead attachment and the low bandwidth of the feedback system (i.e. ∼140 Hz). Although the forces measured are surely underestimates, they are still meaningful for comparative purposes across mutants. The technique of "mean-variance" (MV) analysis (25Patlak J.B. Biophys. J. 1993; 65: 29-42Abstract Full Text PDF PubMed Scopus (89) Google Scholar) was used to derive estimates of force and displacement from the laser trap data. MV analysis begins with a model-independent transformation of the time record, thus giving an alternative view of the data (a MV histogram), which emphasizes intervals of constant properties within the data. Generation of the MV histogram requires no assumptions about or interpretation of the underlying data, and quantitative descriptions of the data are derived from curve fits to the histogram. Thus, MV analysis is less prone to the biases introduced by manual scoring methods and may be used to estimate the size (i.e. d and F), distribution, number, and duration of events (t on) in the data as described previously (7Guilford W.H. Dupuis D.E. Kennedy G. Wu J. Patlak J.B. Warshaw D.M. Biophys. J. 1997; 72: 1006-1021Abstract Full Text PDF PubMed Scopus (223) Google Scholar, 24Lauzon A.-M. Tyska M.J. Rovner A.S. Freyzon Y. Warshaw D.M. Trybus K.M. J. Muscle Res. Cell Motil. 1998; 19: 825-837Crossref PubMed Scopus (119) Google Scholar). Two double-headed constructs with neck lengths shorter than that of wild type HMM were engineered and expressed in the baculovirus/insect cell system (Fig. 1). One construct lacked an RLC binding site (-R site), and SDS-gels confirmed that this construct did not bind RLC (Fig. 2 C). The second shorter necked construct lacked both an RLC and an ELC binding site (neckless). Metal-shadowed images showed that the two globular domains closely abut the rod, as would be expected from a neckless construct that lacked the light chain binding domain (Fig.2 A). These characterizations, as well as the functional assays that follow, are necessary to ensure that the constructs that are analyzed in the laser trap retain certain essential and characteristic features of myosin. To assess if the mutations altered the kinetics of the cross-bridge cycle, actin-activated ATPase measurements were performed. Since both shorter necked constructs lacked the RLC, these molecules did not require phosphorylation for activation. The actin-activated ATPase activity of -R site (open squares) and neckless (filled triangles) was similar to that of phosphorylated WT HMM (filled circles) (Fig.3 A). The mutants were also tested by an in vitro motility assay, which serves as a simplified model system to assess the motion-generating capacity of myosin at the molecular level. Both shorter necked constructs moved actin more slowly than WT HMM. The neckless construct moved actin at ∼25% of the velocity (0.26 ± 0.04 μm/s) of phosphorylated WT HMM (1.1 ± 0.19 μm/s), while -R site moved actin at ∼45% the rate (0.49 ± 0.18 μm/s) of phosphorylated WT HMM (Fig.4 A). To understand the molecular basis for the differences in actin filament velocities between the various constructs, the mechanical and kinetic properties of these constructs were characterized in the laser trap. At the single molecule level, actin filament velocity (v max) is related tod, the unitary displacement generated by myosin during the powerstroke, and t on, the time spent attached to actin following the powerstroke by the equationv max ≅d/t on. Estimates of d andt on were obtained for the mutants and compared with WT HMM by analyzing displacement time series data obtained in the laser trap assay (Fig. 5). Displacement records were characterized by periods of Brownian noise in which HMM-generated displacement events were interspersed. The Brownian noise is reduced upon attachment of HMM to actin (5Finer J.T. Simmons R.M. Spudich J.A. Nature. 1994; 368: 113-119Crossref PubMed Scopus (1583) Google Scholar, 6Molloy J.E. Burns J.E. Kendrick-Jones J. Tregear R.T. White D.C.S. Nature. 1995; 378: 209-212Crossref PubMed Scopus (527) Google Scholar, 7Guilford W.H. Dupuis D.E. Kennedy G. Wu J. Patlak J.B. Warshaw D.M. Biophys. J. 1997; 72: 1006-1021Abstract Full Text PDF PubMed Scopus (223) Google Scholar), which serves as a means of identifying events through the mean-variance analysis technique (see "Experimental Procedures"). This approach becomes a critical tool particularly for analyzing records from shorter necked mutants, which generate displacement events that are small in amplitude and well within the Brownian noise of the system. Based on this analysis, neckless generated an average displacement of 2.0 nm, significantly smaller than the 10.5-nm displacement obtained with WT HMM. The -R site mutant, which has half the neck length of WT HMM, also produced smaller displacements (d = 6.2 nm) than WT HMM (Figs. 5 and 6 and TableI). The d values obtained here for WT HMM are similar to those previously reported for smooth muscle myosin as well as expressed WT HMM under similar conditions (7Guilford W.H. Dupuis D.E. Kennedy G. Wu J. Patlak J.B. Warshaw D.M. Biophys. J. 1997; 72: 1006-1021Abstract Full Text PDF PubMed Scopus (223) Google Scholar, 24Lauzon A.-M. Tyska M.J. Rovner A.S. Freyzon Y. Warshaw D.M. Trybus K.M. J. Muscle Res. Cell Motil. 1998; 19: 825-837Crossref PubMed Scopus (119) Google Scholar,26Tyska M.J. Dupuis D.E. Guilford W.H. Patlak J.B. Waller G.S. Trybus K.M. Warshaw D.M. Lowey S. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4402-4407Crossref PubMed Scopus (155) Google Scholar). Analyses of mean displacement event durations,t on, showed that the -R site mutant had shorter event durations than WT HMM (Table I).Table ISummary of mechanical dataConstructMotility,v maxLaser trapdt onnFnμm/snmmspNShort neck constructs Control WT HMM1.1 ± 0.19 (4)10.5 ± 0.8196 ± 5051.6 ± 0.35 Short necks Neckless0.26 ± 0.04 (3)2.0 ± 0.6 1-aSignificantly different from WT HMM.ND51.1 ± 0.72 -R site0.49 ± 0.18 (2)6.2 ± 1.2 1-aSignificantly different from WT HMM.45 ± 66NDLong neck constructs Control CABL-HMM0.45 ± 0.08 (4)9.1 ± 0.864 ± 5100.9 ± 0.43 Long necks CABL-giraffe0.42 ± 0.08 (4)12.4 ± 0.6 1-bSignificantly different from CABL-HMM.85 ± 6142.1 ± 0.44Values for v max are expressed as the mean ± S.D. with the number in parentheses equaling the number of independent protein preparations. The values for all laser trap parameters are expressed as the mean ± S.E. For these data, nindicates the number of independent data sets analyzed by MV analysis (see "Experimental Procedures"). ND, not determined.1-a Significantly different from WT HMM.1-b Significantly different from CABL-HMM. Open table in a new tab Values for v max are expressed as the mean ± S.D. with the number in parentheses equaling the number of independent protein preparations. The values for all laser trap parameters are expressed as the mean ± S.E. For these data, nindicates the number of independent data sets analyzed by MV analysis (see "Experimental Procedures"). ND, not determined. Unitary forces were also recorded (Fig. 5, Table I). The appearance of force events within the time series data was similar to that of displacement events and thus analyzed by MV analysis. The neckless construct generated forces that were somewhat lower than that of WT HMM (∼1.1 versus ∼1.6 piconewtons, respectively). If the lever arm model appropriately describes the mechanical properties of the neck region, then a longer neck should result in larger unitary displacements. Therefore, a longer necked "giraffe" construct with an additional ELC binding site was expressed. This type of construct is a more stringent test of the lever arm hypothesis (Fig. 1), given that a gain rather than a loss in function is expected. Visual inspection of giraffe HMM by metal-shadowing showed that the construct appeared normal except for having an apparently longer neck region (Fig. 2 A). In addition, the construct decorated actin in a regular manner, suggesting that its actin-binding properties were intact (Fig. 2 B). The ratio of ELC/RLC was 1.7–1.9 times larger in the giraffe construct compared with WT HMM (average slope from four gel loadings, two independent preparations), consistent within experimental error with the construct containing an additional ELC binding site. However, despite having actin-activated ATPase activity, less than half of the actin filaments moved in the motility assay, and they moved at a rate at least 2-fold slower than phosphorylated WT HMM. The rates of motility were 0.49 ± 0.13 μm/s for phosphorylated giraffe HMM and 0.23 ± 0.04 μm/s for unphosphorylated giraffe HMM. The addition of exogenous ELC to the preparation had no effect on motility. When three independent preparations of the phosphorylated giraffe HMM were used in an attempt to obtain unitary displacement data, the data were of insufficient quality to perform the MV analysis. Given that smooth muscle myosin's ability to move actin is strictly dependent on light chain phosphorylation, we hypothesized that light chain phosphorylation was not able to fully activate the giraffe HMM construct. If this is true, then a constitutively active mutant that contains the same elongated neck region as the "wild type" giraffe construct should circumvent the problem of poor motility. We had previously shown that mutation of the actin-binding loop to the sequence found in cardiac myosin produced a constitutively active molecule (16Rovner A.S. Freyzon Y. Trybus K.M. J. Biol. Chem. 1995; 270: 30260-30263Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar), and thus we expressed a long necked mutant, CABL-giraffe, that was active even when unphosphorylated (see "Experimental Procedures"). CABL-HMM with a native neck served as the control. Both constructs were analyzed in the unphosphorylated state. The actin-activated ATPase of CABL-giraffe (filled triangles) (V max = 6.7 ± 0.6 s−1, K m = 6 ± 3 μm) was slightly more than twice than of CABL-HMM (filled circles) (V max = 2.9 ± 0.3 s−1, K m = 16 ± 5 μm) (Fig. 3 B). This result shows that the kinetics of the interaction with actin have been altered by the mutation in the neck. When four independent preparations of CABL-giraffe were compared in parallel with CABL-HMM, there were no significant differences in motility, as would have been predicted from a simple lever arm model (Fig. 4 B). One pair of preparations (Fig. 4 B,1) was also analyzed by determining the velocities during 1-s intervals instead of obtaining an average velocity for a longer run, but still no significant differences emerged (Fig. 4 C). This similarity in rates held true under standard motility conditions (25 mm KCl) as well as at higher salt concentrations (60 mm KCl). As previously reported, unphosphorylated CABL-HMM moved actin at ∼50% the rate of phosphorylated WT HMM, and phosphorylation of CABL-HMM had only a slight effect, increasing this value to ∼75% that of phosphorylated WT HMM (16Rovner A.S. Freyzon Y. Trybus K.M. J. Biol. Chem. 1995; 270: 30260-30263Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). A second long necked construct, CABL-giraffe's twisted sister (see "Experimental Procedures") was expressed in order to test if the relative orientation of the two ELCs had an impact on the rate of motility. The motility observed with this construct was 0.54 ± 0.07 μm/s (number of filaments = 26). This is within the range of values observed for both CABL-HMM and the original CABL-giraffe construct (0.36–0.55 μm/s; see Fig. 4). Despite the fact that CABL-giraffe did not show an increased velocity in the in vitro motility assay (Table I), it is still possible that its unitary displacements could be larger. This in fact proved to be the case. CABL-giraffe generated an average displacement of 12.4 nm, 1.4-fold greater than its control, CABL-HMM, which produced 9.1-nm steps. The mutation at the actin binding loop did not affect step size per se, since the 9.1-nm step of CABL-HMM is statistically indistinguishable from the 10.5-nm unitary step size of WT HMM. The t on for CABL-giraffe was marginally longer than for its control CABL-HMM, but both CABL constructs have significantly shorter values than obtained for WT HMM. Unitary forces produced by CABL-giraffe were approximately twice that generated by CABL-HMM (Table I). Given the higher force-generating capacity of the CABL-giraffe, it is unlikely that the addition of an ELC site introduced a significant compliance within the neck, which potentially could have explained why the rate of motility of CABL-giraffe was not faster than its control CABL-HMM. The goal of this study was to provide a stringent test for whether the myosin neck region acts as a mechanical lever that transmits force and displacements originating within the motor domain. We therefore characterized the molecular mechanics of smooth muscle HMM mutants with shorter and longer neck regions by measuring in vitromotility velocities, unitary displacements, and unitary forces. If the neck acts as a simple lever arm, one prediction is that under unloaded conditions in the laser trap, the magnitude of d should be directly related to lever arm length. Unitary displacements of two short necked constructs were 20% (neckless) and 60% (-R site) of the values obtained with WT HMM. Strikingly, the long necked giraffe mutant generated displacements that were 1.4 times that of its control. The relationship between unitary displacements and lever arm length is linear, strongly supporting the hypothesis that the neck acts as a lever arm (Fig. 6). Another conclusion of this study is that the lever must extend into the motor domain, since our neckless mutant generated significant force and motion. Motion was also observed for other neckless myosin species created genetically in Dictyostelium myosin (27Itakura S. Yamakawa H. Toyoshima Y.Y. Ishijima A. Kojima T. Harada Y. Yanagida T. Wakabayashi T. Sutoh K. Biochem. Biophys. Res. Commun. 1993; 196: 1504-1510Crossref PubMed Scopus (107) Google Scholar) or proteolytically from skeletal muscle myosin (28Waller G.S Ouyang G. Swafford J. Vibert P. Lowey S. J. Biol. Chem. 1995; 270: 15348-15352Abstract Full Text Full Text PDF PubMed Scopus (48) Google Scholar). The portion of the long α-helix that remains in the neckless construct (residues 778–790), which abuts the more compact domain of the converter region (residues 721–777), could provide an additional piece of lever arm for generating force and motion in this construct. The crystal structures of skeletal subfragment 1 and smooth muscle MDE show that the rotation of the converter and the lever arm occurs because of two perpendicular rotations around two conserved glycines that are located at either end of the SH1 helix (4Dominguez R. Freyzon Y. Trybus K.M. Cohen C. Cell. 1998; 94: 559-571Abstract Full Text Full Text PDF PubMed Scopus (591) Google Scholar). In agreement with the structural data, the correlation between neck length and d (Fig. 6) suggests that the fulcrum for movement is ∼3 nm within the motor domain. Therefore, the bulk of the motor domain remains in a relatively fixed orientation with respect to actin, and the point at which myosin undergoes a major rotation is closer to the converter region than to the actomyosin interface. The lever hypothesis also makes specific predictions about the relationship between lever arm length (L) and unitary force (see Fig. 7), but these are model-dependent. In contrast, all three models described below predict a linear relationship between lever arm length and unitary displacement (Fig. 6). Our data showed that neckless generated somewhat less force than WT HMM, while the giraffe construct generated more force than its control. These data rule out the simplest lever arm model, which would have force inversely related to L. A second model, suggested by Howard and Spudich (Appendix in 13), proposed that the neck region acts more like an elastic cantilever rather than a rigid lever and predicts that force (F) should be inversely related to L 2. Our data also do not fit this model, which was recently used to explain changes in x-ray diffraction intensity data with small changes in muscle length (29Dobbie I. Linari M. Piazzesi G. Reconditi M. Koubassova N. Ferenczi M.A. Lombardi V. Irving M. Nature. 1998; 396: 383-387Crossref PubMed Scopus (133) Google Scholar), where the intensity changes were assumed to be the result of both a rotation and bending of the neck region. A third mechanical model, that we previously proposed to explain the force- and motion-generating capacity of light chain-deficient skeletal muscle myosins (10Lowey S. Waller G.S. Trybus K.M. Nature. 1993; 365: 454-456Crossref PubMed Scopus (268) Google Scholar, 30VanBuren P. Waller G.S. Harris D.E. Trybus K.M. Warshaw D.M. Lowey S. Proc. Natl. Acad. Sci. 1994; 91: 12403-12407Crossref PubMed Scopus (122) Google Scholar), predicts that force should be linearly related to L. In this model, a torque generated in the motor domain produces force at the end of the neck, which acts as a rigid lever. This force would then stretch an elastic element, presumably in the subfragment 2 region (30VanBuren P. Waller G.S. Harris D.E. Trybus K.M. Warshaw D.M. Lowey S. Proc. Natl. Acad. Sci. 1994; 91: 12403-12407Crossref PubMed Scopus (122) Google Scholar). A comparison of the force data from the laser trap with the predictions for the three models (see Fig. 7) suggests that the third model may be the most appropriate. Both the displacement and the force data support the conclusion that the neck region acts as a rigid lever and that much of the myosin elasticity is external to the neck. It should be noted that force measurements in this assay are underestimates, given the stray compliances within the experimental system (see "Experimental Procedures"), and thus conclusions based on these data should be viewed in light of this concern. An alternative view of the neck has been championed by Yanagida and co-workers (31Tanaka H. Kitamura K. Iwane E.H. Yanagida T. Biophys. J. 2000; 78 (abstr.): 3AGoogle Scholar). Although they also obtained a reduced velocity with a similar neckless construct of Dictyostelium myosin, they showed that the unitary step size was unaffected by the change in neck length and therefore concluded that the slower velocity was due only to an increase in the time of attachment following the powerstroke (31Tanaka H. Kitamura K. Iwane E.H. Yanagida T. Biophys. J. 2000; 78 (abstr.): 3AGoogle Scholar). This view favors the idea that the neck region serves to regulate the kinetics of the cross-bridge cycle. In contrast, our unitary displacement data support the idea that the neck acts as a mechanical lever arm. However, we also have evidence that kinetic changes have occurred with several of the neck length constructs, both in the unitary event durations (Table I) and in steady-state actin-activated ATPase activity (Fig. 3 B). Thus, the assumption thatton remains constant and that all changes in velocity are due to changes in unitary displacement (v max ≅d/t on), which was made in several previous studies with altered neck length mutants (12Uyeda T.Q.P. Spudich J.A. Science. 1993; 262: 1867-1870Crossref PubMed Scopus (103) Google Scholar, 13Uyeda T.Q. Abramson P.D. Spudich J.A. Proc. Natl. Acad. Sci. 1996; 93: 4459-4464Crossref PubMed Scopus (392) Google Scholar, 14Anson M Geeves M.A. Kurzawa S.E. Manstein D.J. EMBO J. 1996; 15: 6069-6074Crossref PubMed Scopus (136) Google Scholar), is probably not justified. For example, CABL-giraffe moved actin at the same rate as its control, CABL-HMM, although the step size was increased, in contrast to a previous study, which reported that a long necked myosin showed faster motility than WT myosin (13Uyeda T.Q. Abramson P.D. Spudich J.A. Proc. Natl. Acad. Sci. 1996; 93: 4459-4464Crossref PubMed Scopus (392) Google Scholar). A kinetic contribution to changes in velocity because of neck alterations has also been inferred from several previous studies. Changes in the unloaded duty cycle contributed to the reduced velocity and force production of ELC-deficient myosin (30VanBuren P. Waller G.S. Harris D.E. Trybus K.M. Warshaw D.M. Lowey S. Proc. Natl. Acad. Sci. 1994; 91: 12403-12407Crossref PubMed Scopus (122) Google Scholar). Point mutations in the RLC have been shown to affect in vitro motility rates although ATPase activity remains high (32Chaudoir B.M. Kowalczyk P.A. Chisholm R.L. J. Cell Sci. 1999; 112: 1611-1620PubMed Google Scholar). Finally, a RLC point mutant, with weakened divalent cation binding, modulates the kinetics of cross-bridge attachment and detachment (33Diffee G.M. Patel J.R. Reinach F.C. Greaser M.L. Moss R.L. Biophys. J. 1996; 71: 341-350Abstract Full Text PDF PubMed Scopus (35) Google Scholar). Assuming that the unitary step size was unchanged by this point mutation, these data also argue for the potential modulation of velocity via kinetic mechanisms as a result of structural perturbations to the neck. We conclude that changes in the length of the light chain binding domain affect the unitary displacement and force of both shorter and longer necked constructs in a way that is consistent with the neck acting as a lever arm. The fulcrum for the rotation is located ∼3 nm internal to the C-terminal end of the motor domain, far from the actomyosin interface. Perturbations to the neck region also appear to modulate the kinetics of several constructs, as judged by changes in steady state ATPase activity and by changes in the unitary displacement event durations. It is possible that kinetic changes are accentuated in these smooth muscle myosin constructs, where myosin activity is tightly regulated by light chain phosphorylation. Thus, the neck region acts as a lever arm but may also serve to transmit mechanical strain to the catalytic site within the motor domain. We thank Eric Hayes, Janet Vose, and Greta Ouyang for technical assistance, Anne-Marie Lauzon for assisting in gathering data early on in the study, Don Gaffney for programming expertise, and Guy Kennedy for instrumentation skills that were needed to keep the trap functioning throughout this study.