Advertisement

Functional Characterization of the N-terminal Region of Myosin-2*

  • Setsuko Fujita-Becker
    Affiliations
    Max-Planck-Institut für Medizinische Forschung, Jahnstrasse 29, 69120 Heidelberg, Germany
    Search for articles by this author
  • Georgios Tsiavaliaris
    Affiliations
    Institut für Biophysikalische Chemie, Medizinische Hochschule Hannover, Carl-Neubergstrasse 1, 30623 Hannover, Germany
    Search for articles by this author
  • Reiko Ohkura
    Affiliations
    Department of Life Science, University of Tokyo, Komaba 3-8-1, Tokyo 153-8902, Japan
    Search for articles by this author
  • Takashi Shimada
    Affiliations
    Department of Life Science, University of Tokyo, Komaba 3-8-1, Tokyo 153-8902, Japan
    Search for articles by this author
  • Dietmar J. Manstein
    Correspondence
    To whom correspondence should be addressed: Medizinische Hochschule Hannover, Institute for Biophysical Chemistry, OE4350, 30625 Hannover, Germany. Tel.: 49-511-5323700; Fax: 49-511-532-5966;
    Affiliations
    Max-Planck-Institut für Medizinische Forschung, Jahnstrasse 29, 69120 Heidelberg, Germany

    Institut für Biophysikalische Chemie, Medizinische Hochschule Hannover, Carl-Neubergstrasse 1, 30623 Hannover, Germany
    Search for articles by this author
  • Kazuo Sutoh
    Affiliations
    Department of Life Science, University of Tokyo, Komaba 3-8-1, Tokyo 153-8902, Japan
    Search for articles by this author
  • Author Footnotes
    * This work was supported by Grant MA1081/5-3 from the Deutsche Forschungsgemeinschaft. 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.
      All class 2 myosins contain an N-terminal extension of ∼80 residues that includes an Src homology 3 (SH3)-like subdomain. To explore the functional importance of this region, which is also present in most other myosin classes, we generated truncated constructs of Dictyostelium discoideum myosin-2. Truncation at position 80 resulted in the complete loss of myosin-2 function in vivo. Actin affinity was more than 80-fold, and the rate of ADP release ∼40-fold decreased in this mutant. In contrast, a myosin construct that lacks only the SH3-like subdomain, corresponding to residues 33-79, displayed much smaller functional defects. In complementation experiments with myosin-2 null cells, this construct rescued myosin-2-dependent processes such as cytokinesis, fruiting body formation, and sporogenesis. An 8-fold reduction in motile activity and changes of similar extent in the affinity for ADP and filamentous actin indicate the importance of the SH3-like subdomain for correct communication between the functional regions within the myosin motor domain and suggest that local perturbations in this region can play a role in modulating myosin-2 motor activity.
      Members of the myosin superfamily of actin-based motors act in a variety of cellular functions such as muscle contraction, cell and organelle movement, membrane trafficking, and signal transduction. Although myosin motor domains show a high degree of sequence conservation, the individual myosin classes are clearly defined by differences in the head structure (
      • Sellers J.R.
      ). Extensive biochemical investigations of the myosin ATPase cycle together with structural information of the motor domain and electron microscopy of the actomyosin complex have led to detailed molecular models of the nucleotide-dependent actomyosin interaction (
      • Holmes K.C.
      • Schröder R.R.
      • Sweeney H.L.
      • Houdusse A.
      ,
      • Geeves M.A.
      • Fedorov R.
      • Manstein D.J.
      ). However, the exact functional roles of several regions of the myosin motor domain remain to be elucidated. In particular, the role of a protruding, six-stranded, antiparallel, β-barrel subdomain with similarities to the SH3
      The abbreviations used are: SH3, Src homology 3; ELC, myosin-2 essential light chain; M761, D. discoideum myosin-2 motor domain; ΔN1-M761, D. discoideum myosin-2 motor domain lacking the 80 N-terminal residues (ΔN2 constructs have residues Y34-N78 replaced with the tripeptide GTG); F-actin, filamentous actin; Ni2+-NTA, nickel-nitrilotriacetic acid; DTT, dithiothreitol; MOPS, 4-morpholinepropanesulfonic acid; PIPES, 1,4-piperazinediethanesulfonic acid; pyr, pyrene; M, myosin.
      2The abbreviations used are: SH3, Src homology 3; ELC, myosin-2 essential light chain; M761, D. discoideum myosin-2 motor domain; ΔN1-M761, D. discoideum myosin-2 motor domain lacking the 80 N-terminal residues (ΔN2 constructs have residues Y34-N78 replaced with the tripeptide GTG); F-actin, filamentous actin; Ni2+-NTA, nickel-nitrilotriacetic acid; DTT, dithiothreitol; MOPS, 4-morpholinepropanesulfonic acid; PIPES, 1,4-piperazinediethanesulfonic acid; pyr, pyrene; M, myosin.
      domains, which appears to be present in most myosins and comprises ∼50 amino acids of the heavy chain, is largely unknown. The structure of this subdomain has been solved for class 2, 5, and 6 myosins (
      • Menetrey J.
      • Bahloul A.
      • Wells A.L.
      • Yengo C.M.
      • Morris C.A.
      • Sweeney H.L.
      • Houdusse A.
      ,
      • Coureux P.D.
      • Wells A.L.
      • Menetrey J.
      • Yengo C.M.
      • Morris C.A.
      • Sweeney H.L.
      • Houdusse A.
      ,
      • Smith C.A.
      • Rayment I.
      ,
      • Schroder R.R.
      • Manstein D.J.
      • Jahn W.
      • Holden H.
      • Rayment I.
      • Holmes K.C.
      • Spudich J.A.
      ). The crystal structure of the nucleotide-free smooth muscle myosin motor domain with essential light chain (ELC) bound, shows the ELC to be in contact with the N-terminal domain of the heavy chain. However, in the presence of MgADP·AlF4- the ELC is rotated up to 70° from the position in nucleotide-free myosin subfragment 1 (S1) and forms a contact with loop 1 (
      • Dominguez R.
      • Freyzon Y.
      • Trybus K.M.
      • Cohen C.
      ).
      Sequence alignments of the N-terminal region of myosins from different classes reveal that this region varies greatly in length and amino acid composition among the individual members. Class 1 myosins completely lack the N-terminal region corresponding to the first 79 residues of myosin-2 (
      • Kollmar M.
      • Durrwang U.
      • Kliche W.
      • Manstein D.J.
      • Kull F.J.
      ). In the case of class 2 myosins, residues 1-33 form an extended structure crossing the interface between the motor domain and the neck region (Fig. 1). Residues from this segment are in close contact and cover a hydrophobic groove on the motor domain surface. The ensuing residues 33-79 form the SH3-like subdomain (
      • Koch C.A.
      • Anderson D.
      • Moran M.F.
      • Ellis C.
      • Pawson T.
      ). Biochemical studies with zero-length cross linkers indicate that the two heads of myosin, when bound in a rigor complex with F-actin, are in close contact with each other. In the case of chicken gizzard HMM (heavy meromyosin), Glu168 could be cross-linked to Lys65 in the SH3-like subdomain of the neighboring head (
      • Onishi H.
      • Maita T.
      • Matsuda G.
      • Fujiwara K.
      ). The cross-linked acto-HMM displayed a significantly increased ATPase activity, which was interpreted as evidence that this interaction may play a pivotal role in myosin function (
      • Schroder R.R.
      • Manstein D.J.
      • Jahn W.
      • Holden H.
      • Rayment I.
      • Holmes K.C.
      • Spudich J.A.
      ,
      • Ito K.
      • Liu X.
      • Katayama E.
      • Uyeda T.Q.
      ).
      Figure thumbnail gr1
      FIGURE 1Structure of the myosin-2 motor domain and its N-terminal region. A, ribbon representation of the Dictyostelium myosin-2 motor domain (
      • Kollmar M.
      • Durrwang U.
      • Kliche W.
      • Manstein D.J.
      • Kull F.J.
      ). Residues 81-689 of the motor domain are shown in green. The N-terminal region corresponding to residues 2-33 is shown in yellow. The SH3-like subdomain formed by residues 34-78 (yellow) was truncated in constructs ΔN2-M761 and ΔN2-myosin. The entire N-terminal region formed by residues 2-80 was truncated in constructs ΔN1-M761 and in the equivalent full-length construct ΔN1-myosin. The converter region and the C-terminal α-helix emerging from the motor domain (residues 690-770) are shown in blue. Hydrophobic interactions formed between residues 760-765 in this helix and residues Phe25-Lys32 provide additional structural stability. B, hydrophobic interactions formed between the N-terminal region and the core motor domain. Ribbon representation and electrostatic surface potential of the 80-amino acid-long N-terminal region of Dictyostelium myosin-2. Negatively charged residues are labeled in red, noncharged residues in white, and positively charged residues in blue. Hydrophobic side chains Pro3, Ile4, Tyr11, Leu15, and Tyr14 (yellow), shown as a ball-and-stick model, are part of a disordered loop (red) facing a groove formed by hydrophobic residues Pro133, Ile134, Met139, Ile142, Phe143, and His154 (white surfaces). C, the view from B rotated by 90°. D, amino acid sequence of the Dictyostelium myosin-2 N terminus. The N-terminal region of wild-type myosin-2 (WT) comprises the first 80 amino acids. This region has been removed in ΔN1-myosin and the corresponding motor domain construct ΔN1-M761. In the ΔN2-constructs only the SH3-like domain (residues Tyr34-Asn78) has been replaced by the tripeptide Gly-Thr-Gly. This figure was prepared using the programs WebLab Viewer Lite 4.0 (Molecular Simulations Inc.) and POV-Ray 3.6 (Persistence of Vision Ray Tracer v3.02).
      Here, we have focused on the biological and biochemical properties of the N-terminal region of myosin-2. We used deletion mutants to elucidate the role of this region by analyzing the kinetic behavior of the mutants and their ability to produce force and movement in an in vitro assay system. Complementation of Dictyostelium discoideum myosin-2 null cells was used for the in vivo analysis of the N-terminally truncated constructs. The characterization of the truncated myosins demonstrates that the removal of the SH3-like domain affects the interaction with nucleotides and actin as well as communication within the motor domain. Complete removal of the 80-amino acid-long N-terminal region resulted in almost complete loss of function and highly aberrant kinetic properties.

      EXPERIMENTAL PROCEDURES

      Plasmid Construction and Transformation—Standard chemicals were purchased from Sigma, and restriction enzymes were obtained from MBI-Fermentas (St. Leon-Roth, Germany) and New England Biolabs (Frankfurt, Germany). Escherichia coli strain XL1Blue (Stratagene, Heidelberg, Germany) was used for amplification of plasmids. All cloning was performed using standard procedures. The expression vectors for the production of the individual myosin constructs were based on the extrachromosomal vector pDXA-3H (
      • Manstein D.J.
      • Schuster H.P.
      • Morandini P.
      • Hunt D.M.
      ). Two types of N-terminally modified Dictyostelium myosin-2 constructs were generated: ΔN1 constructs have the first N-terminal 80 amino acids deleted, and ΔN2 constructs have residues Tyr34 to Asn78, corresponding to the SH3-like domain, replaced with the tripeptide GTG (Fig. 1).
      Plasmid pSA1 encodes for the expression of full-length Dictyostelium myosin-2 and carries the sequence of a C-terminal His8 tag as described previously (
      • Manstein D.J.
      • Hunt D.M.
      ). PCR-directed mutagenesis using pSA1 as template resulted in the generation of myosin constructs pΔN-1-myosin and pΔN2-myosin encoding full-length myosin-2 with a completely or partially truncated N terminus, respectively. In pΔN2-myosin the base triplets encoding Tyr34 to Asn78 were deleted and replaced with the sequence GGTACCGGT to introduce a KpnI-AgeI site in the pSA1 plasmid. The N-terminal sequence was changed to introduce a BamHI site. Accordingly, the N terminus of ΔN2-myosin starts with Met-Asp-Pro, and residues Tyr34 to Asn78 are replaced by the tripeptide Gly-Thr-Gly (Fig. 1). Subcloning of the 2-kb SalI-BstXI gene fragments from pΔN-1-myosin or pΔN2-myosin into pM761 (
      • Kurzawa S.E.
      • Manstein D.J.
      • Geeves M.A.
      ) produced constructs pΔN-1-M761 and pΔN2-M761, respectively, encoding the myosin-2 motor domain with truncated N terminus. Dictyostelium AX3-ORF+ cells were used for the production of the motor domain constructs. Wild-type myosin and N-terminally truncated full-length myosins ΔN1-myosin and ΔN2-myosin were produced in the mhcA- null cell line, HS1 (
      • Manstein D.J.
      • Titus M.A.
      • De Lozanne A.
      • Spudich J.A.
      ). Transformations were performed by electroporation (
      • Egelhoff T.T.
      • Titus M.A.
      • Manstein D.J.
      • Ruppel K.M.
      • Spudich J.A.
      ). Transformants were screened for protein production as described previously (
      • Manstein D.J.
      • Hunt D.M.
      ). Cell lines producing near wild-type levels of the recombinant myosins were selected for further analysis.
      Protein Expression and PurificationDictyostelium AX3-ORF+ cells producing the motor domain constructs and ΔN2-myosin myosins were grown as described previously (
      • Manstein D.J.
      • Hunt D.M.
      ). Cells producing ΔN1-myosin myosin were grown on 26 × 26-cm plastic plates filled with 100 ml of HL-5C medium. The confluent plates were incubated additionally for 24 h on gyratory shakers at 40 rpm before harvesting. Myosin null cells (HS1) transformed with pSA1, pΔN1-myosin, or pΔN2-myosin produced levels of full-length myosin and N-terminally truncated myosins similar to that of wild-type AX3-ORF+ cells (data not shown). The differences in molecular weight between the myosins were determined by SDS-PAGE using a 4-12% gradient (data not shown). The motor domain constructs were purified by Ni2+-NTA chromatography giving yields of 4 mg/g of cells for M761, 0.5 mg/g of cells for ΔN1-M761, and 1 mg/g of cells for ΔN2-M761. Some minor modifications were made for the purification of ΔN1-M761. Because of its lower thermal stability and increased propensity to aggregate, ΔN1-M761 was purified in the presence of 100 mm KCl, and centrifugation was performed at 30,000 × g. All buffers contained 100 mm KCl, and centrifugation steps were changed from 230,000 × g for 1 h to 25,000 × g for 30 min and from 500,000 × g for 1 h to 75,000 g for 30 min. Purified ΔN1-M761 was concentrated by dialysis against solid sucrose. The ATPase activity of ΔN1-M761 was greatly reduced after frozen storage of the protein. Therefore, we determined the time-dependent reduction of the actin-activated ATPase activity of the protein upon storage on ice. This showed that ΔN1-M761 displayed no significant reduction in enzymatic activity during the first 48 h. The results shown here were obtained with ΔN1-M761 from five different preparations, and all measurements were performed within 24 h after elution of the protein from the Ni2+-NTA column. The results shown for ΔN2-M761 and the myosin constructs with a complete tail region are based on at least three separate preparations.
      Following purification by Ni2+-NTA affinity chromatography (
      • Manstein D.J.
      • Hunt D.M.
      ), yields of 0.5, 1.5, and 4.0 mg of purified protein were obtained for ΔN1-M761, ΔN2-M761, and M761 per gram of cells. The reduced yields that were obtained with ΔN1-M761 and ΔN2-M761 result from increased losses during the initial steps of the purification including the wash step prior to ATP extraction of the recombinant motor domains.
      Full-length myosins were prepared by the method of Ruppel et al. (
      • Ruppel K.M.
      • Uyeda T.Q.
      • Spudich J.A.
      ) with some modifications. After lysis, centrifugation, and washing, the myosins were extracted from pellets with extraction buffer containing 10 mm HEPES, pH 7.4, 125 mm NaCl, 3 mm MgCl2, 1 mm DTT, and 3 mm ATP. After dialyses against the buffer containing 10 mm PIPES, pH 6.8, 50 mm NaCl, 10 mm MgCl2, and 1 mm DTT, the precipitated myosins were resolved in extraction buffer containing 300 mm NaCl. The assembly-disassembly cycle was repeated, and myosin was finally dissolved in 0.2 volume/g cell of storage buffer containing 10 mm HEPES, pH 7.4, 250 mm NaCl, 1 mm DTT, 3 mm MgCl2, and 2 mm ATP. The purified myosins were treated with Dictyostelium myosin light chain kinase as described by Ruppel et al. (
      • Ruppel K.M.
      • Uyeda T.Q.
      • Spudich J.A.
      ). Rabbit actin was purified and labeled with pyrene (pyrene-actin) as described previously (
      • Criddle A.H.
      • Geeves M.A.
      • Jeffries T.
      ). The concentrations of the recombinant Dictyostelium myosin motor domains were determined using the Bradford assay.
      Direct Functional Assays—Actin-sliding motility was performed at 30 °C using a fluorescence microscope as described previously (
      • Anson M.
      • Geeves M.A.
      • Kurzawa S.E.
      • Manstein D.J.
      ) with some minor modifications. Assay buffer contained 10 mm DTT. Experimental flow cells were constructed using bovine serum albumin (0.5 mg/ml in assay buffer)-coated glass slides and nitrocellulose-coated coverslips. Myosins were actin affinity-purified immediately before use, to remove rigor-forming myosins, as described by Uyeda et al. (
      • Uyeda T.Q.
      • Kron S.J.
      • Spudich J.A.
      ). The sliding movement started by introducing 2 mm ATP into bovine serum albumin/anti-fade solution containing 0.5% methylcellulose.
      Kinetic Measurements—Stopped-flow measurements were performed at 20 °C with a Hi-tech Scientific SF61 or an Applied Photophysics PiStar stopped-flow spectrophotometer using procedures and kinetic models described previously (
      • Batra R.
      • Geeves M.A.
      • Manstein D.J.
      ,
      • Furch M.
      • Fujita-Becker S.
      • Geeves M.A.
      • Holmes K.C.
      • Manstein D.J.
      ,
      • Tsiavaliaris G.
      • Fujita-Becker S.
      • Batra R.
      • Levitsky D.I.
      • Kull F.J.
      • Geeves M.A.
      • Manstein D.J.
      ). All concentrations refer to the concentration of the reactants after mixing in the stopped-flow observation cell. A notation is used that distinguishes between rate and equilibrium constants in the presence and absence of actin by using bold (k+1, K1) versus italic type (k+1, K1); subscripts A and D refer to actin (KA) and ADP (KD), respectively. Steady state ATPase activities were determined at 25 °C using a linked enzyme assay and analyzed as described (
      • Furch M.
      • Geeves M.A.
      • Manstein D.J.
      ). The myosin concentration was 0.5-1 μm, and the highest actin concentration was 60 μm. NADH oxidation was followed using the change in absorption at 340 nm in a Beckman DU-650 spectrophotometer (Beckman-Coulter, Dreeich, Germany). The conditions were: 25 mm imidazole, 25 mm KCl, 4 mm MgCl2, 0.5 mm DTT, 0.5 mm ATP, 0.2 mm NADH, 0.5 mm phosphoenol pyruvate, 0.02 mg/ml lactate dehydrogenase, 0.05 mg/ml pyruvate kinase, pH 7.4. The error bars shown in the graphs represent the standard deviations from at least 15 determinations of each data point. Protein from three or more preparations per construct was used for the generation of the graphs.

      RESULTS

      Kinetic Properties of ΔN1-M761 and ΔN2-M761—To analyze in detail the effects of the truncations on the interaction with nucleotides and F-actin, steady-state ATPase measurements and transient kinetics measurements were performed. For each construct the actin-activated ATPase activity was measured over a range of 0 to 80 μm F-actin. In the presence of saturating concentrations of ATP and in the absence of F-actin, the basal ATPase activity of ΔN1-M761 was 0.014 s-1 and thus ∼2-fold slower than the value determined for M761. The basal ATPase rate of ΔN2-M761 was 0.05 s-1 (Table 1). At concentrations of actin much lower than Kapp, the dependence of the ATPase rate on the concentration of F-actin could be fitted to a straight line. The apparent second order rate constant for F-actin binding Kapp/kcat of the reaction could be determined from the slope of this line. A greater than 30-fold reduction in Kapp/kcat was observed for ΔN1-M761 compared with a less than 2-fold reduction for ΔN2-M761 (Table 1). At saturating actin concentrations, kcat values of 0.4 and 0.9 s-1 were determined for ΔN1-M761 and ΔN2-M761. The corresponding value for M761 is 2.1 s-1.
      TABLE 1Actin activation of ATPase activity The experimental conditions were 25 mm HEPES, pH 7.4, 25 mm KCl, 4 mm MgCl2 at 25 °C. Actin-activated ATPase activity was measured in the presence of rabbit skeletal muscle F-actin.
      Myosin constructkbasalkcat
      Values for kcat and Kapp were calculated from fitting the data to the Michaelis-Menten equation
      Kapp
      Values for kcat and Kapp were calculated from fitting the data to the Michaelis-Menten equation
      kcat/Kapp
      The data at concentrations of actin much lower than Kapp could be fit to a straight line, and the apparent second order rate constant for actin binding kcat/Kapp was determined from the slope of this line
      s–1s–1μmμm–1 s–1
      ΔN1-M7610.014 ± 0.0010.37 ± 0.15>300>0.001
      ΔN2-M7610.05 ± 0.0010.87 ± 0.0845 ± 2.40.019
      M7610.037 ± 0.0122.06 ± 0.2870.4 ± 100.029
      a Values for kcat and Kapp were calculated from fitting the data to the Michaelis-Menten equation
      b The data at concentrations of actin much lower than Kapp could be fit to a straight line, and the apparent second order rate constant for actin binding kcat/Kapp was determined from the slope of this line
      Binding of ATP to myosin motor domains in the absence of actin was monitored by the increase of intrinsic protein fluorescence following the addition of excess ATP and analyzed according to the model shown in Scheme 1, where the asterisks indicate changes in intrinsic protein fluorescence. The amplitude of the fluorescence signal obtained with M761 and ΔN2-M761 was 17% and showed little change in the range from 50 μm to 2 mm ATP. In the case of ΔN1-M761, the addition of less than 250 μm ATP produced more than 5-fold less change in intrinsic protein fluorescence, and at higher ATP concentrations the signal change became too small to allow accurate rate determinations. The observed rate constants (kobs) for the exponential increase in protein fluorescence were linearly dependent on the concentration of ATP up to 100 μm. The apparent second order rate constant for ATP binding to myosin (K1k+2) is defined by the slope of the best-fit line and could be determined for all three constructs. Values of K1k+2 for ATP binding were similar for wild-type myosin M761 and ΔN2-M761 and 3-fold faster for ΔN1-M761 (Table 2). At high ATP concentrations (>2 mm) the observed rate constants for M761 and ΔN2-M761 were saturating, and the dependence on the ATP concentration could be described by a hyperbola, where the maximum value of kobs defines k+2, the rate of the conformational change that follows ATP binding and precedes ATP hydrolysis (Scheme 1).
      TABLE 2Kinetic parameters of myosin interaction with nucleotides The experimental conditions were 20 mm MOPS, pH 7.0, 5 mm MgCl2, 100 mm KCl at 20 °C. The binding and hydrolysis of ATP by D. discoideum myosin head fragments was analyzed in terms of Scheme 1. k+i and k–i are forward and reverse rate constants, and Ki (k+i/ki) is the association equilibrium constant of the ith step of the reaction. KD corresponds to K6K7. Uncertainties represent standard errors in the best fits of the data. ND, not determined.
      NucleotideRate constantM761ΔN1-M761ΔN2-M761
      ATPK1k+2m–1 s–1)0.90 ± 0.043.12 ± 0.100.60 ± 0.05
      k+2 (s–1)28 ± 4ND176 ± 9
      mantATPK1k+2m–1 s–1)1.04 ± 0.021.03 ± 0.020.60 ± 0.01
      mantADPk+Dm–1 s–1)0.36 ± 0.010.89 ± 0.010.45 ± 0.005
      k–D (s–1)1.6 ± 0.0020.04 ± 0.0031.3 ± 0.002
      KDm)4.450.0452.9
      The change observed upon the addition of ADP was too small to measure for all three constructs. Therefore, additional measurements were performed using the fluorescent analogues mantATP and mantADP to measure nucleotide binding (
      • Woodward S.K.
      • Geeves M.A.
      • Manstein D.J.
      ). Binding of mant-nucleotides was determined by monitoring the increase in mant-fluorescence upon the addition of increasing concentrations of the fluorescent nucleotides to the motor domain constructs. In the range of 1 to 25 μm, the observed rate constants (kobs) for the exponential increase in fluorescence were linearly dependent on the concentration of mant-nucleotide. The apparent second order rate constants were determined from the slope of the plotted best-fit line (data not shown). The rate for mantATP binding was similar for all constructs. Similarly, the second order rate constants for mantADP binding (k+D) displayed less than 3-fold differences (Table 2). The rate of mantADP dissociation (k-D) was determined by monitoring the decrease in fluorescence upon displacement of mantADP from the myosin·mantADP complex by the addition of excess ATP. The observed process could be fitted to a single exponential (Fig. 2) with kobs corresponding directly to the dissociation rate k-D (see Scheme 1). Compared with M761 and ΔN2-M761 the rate of mantADP release from ΔN1-M761 was decreased by more than 30-fold (Fig. 2, insert). Therefore, the almost 100-fold increase in ADP affinity displayed by ΔN1-M761 was caused mostly by the slower rate of ADP dissociation.
      Figure thumbnail gr3
      FIGURE 2Rate of mantADP displacement from myosin head fragments. Stopped-flow records of fluorescence decrease upon binding of 200 μm ATP to an 0.5 μm myosin construct premixed with 0.5 μm mantADP. M761 and ΔN2-M761 show similar rates of mantADP displacement of 1.6 and 1.3 s-1, respectively. MantADP release from ΔN-1 is ∼40 times slower (insert). The observed dissociation rate constants (k-D) are summarized in .
      The rate of F-actin binding was measured by following the exponential decrease in pyrene fluorescence observed upon binding of excess pyrene-labeled F-actin to myosin head fragments (
      • Kurzawa S.E.
      • Manstein D.J.
      • Geeves M.A.
      ). The change in pyrene fluorescence could be fitted to a single exponential function. The observed rate constants were linearly dependent upon F-actin concentration over the entire range studied (Fig. 3A). The apparent second order rate constants for pyrene-actin binding (k+A) were obtained from the slopes of the plotted line. Values of 1.2 × 106 m-1 s-1 for M761, 0.19 × 106 m-1 s-1 for ΔN1-M761, and 0.10 × 106 m-1 s-1 for ΔN2-M761 were obtained. The rate of F-actin dissociation (k-A) from the myosin constructs was determined from the rate of fluorescence enhancement observed by displacing pyrene-actin from pyr-acto·M with an excess of unlabeled actin (Fig. 3B). M761 and ΔN2-M761 displayed a less than 2-fold difference in the rate of actin displacement. In contrast, F-actin dissociation from ΔN1-M761 was increased 9-fold (Fig. 3B, insert). The dissociation equilibrium constants for actin binding (KA) defined by the ratio k-Ak+A shows that the actin affinities for ΔN1-M761 and ΔN2-M761 are ∼100- and ∼20-fold lower than the value obtained for M761 (Table 3, Schemes 2 and 3). These differences explain, at least in part, the differences in the yields obtained for the different constructs during purification.
      Figure thumbnail gr4
      FIGURE 3Interaction of pyrene-actin with myosin head fragments. A, dependence of the rate of pyrene-actin binding to myosin head fragments on pyrene-actin concentration. The data were fit to a straight line, the slope of which gives the second order rate constants for binding to actin (k+A). All of the resulting values are summarized in . ▪, M761; ○, ΔN1-M761; □, ΔN2-M761. B, the dissociation rate constants (k-A) for each myosin construct was determined from the rate of fluorescence enhancement that followed the addition of excess unlabeled actin to the pyrene-actin-equilibrated myosin head fragment. The solid line is the best fit to single exponential function giving values for (k-A). ΔN1-M761 displays 10 times faster actin dissociation kinetics than M761 and ΔN2-M761. The dissociation equilibrium constant (KA) for actin binding was calculated from the ratios of k-A and k+A (). The experimental conditions were 20 mm MOPS, 5 mm MgCl2, and 100 mm KCl, pH 7.0, at 20 °C.
      TABLE 3Kinetic parameters of myosin interaction with actin The experimental conditions were 20 mm MOPS, pH 7.0, 5 mm MgCl2, 100 mm KCl at 20 °C. Actomyosin ATPase activity was analyzed in terms of Scheme 2. Coupling between the actin and the nucleotide binding sites was analyzed using the model shown in Scheme 3. In this scheme KA, KAD, KDA, and KD are defined as dissociation equilibrium constants. The presence of actin alters the affinity of ADP for myosin and vice versa. The ADP affinity for MHF is given as KD, and the affinity of myosin for actin is defined as KA.
      NucleotideRate constantM761ΔN1-M761ΔN2-M761
      Nucleotide binding to actomyosin
      ATPK1k+2m–1 s–1)0.23 ± 0.010.59 ± 0.0080.17 ± 0.02
      ADPKADm)210 ± 2No signal15 ± 2
      KAD/KD18 ± 0.75
      The value for KD was determined using mantADP instead of ADP. Uncertainties represent standard errors in the best fits of the data
      Actin binding to myosin
      k+Am–1 s–1)1.22 ± 0.10.19 ± 0.030.10 ± 0.01
      K–A (s–1)0.0035 ± 0.00050.044 ± 0.00050.005 ± 0.0004
      KA (nm)2.923050
      KDA = KA·KAD/KD (nm)52.2250
      a The value for KD was determined using mantADP instead of ADP. Uncertainties represent standard errors in the best fits of the data
      Addition of excess ATP to pyrene-actomyosin complexes results in an exponential increase in pyrene fluorescence as the actin dissociates. Values of kobs were linearly dependent on ATP concentration in the range of 5 to 25 μm (data not shown). The slope of the best-fit line defines the apparent second order rate constant (K1k+2) values of 2.3 × 105 m-1 s-1, 5.9 × 105 m-1 s-1, and 1.7 × 105 m-1 s-1 were obtained for M761, ΔN1-M761, and ΔN2-M761, respectively (Table 3).
      The affinity of ADP for pyr-acto·M was determined from the ADP inhibition of the ATP-induced dissociation of actin from the complex. Mixing of 0.25 m pyr-acto·M with 100 μm ATP in the presence of different amounts of ADP resulted in an exponential increase in intrinsic protein fluorescence. No signal was obtained with ΔN1-M761. An increase in the concentration of ADP produced a reduction in kobs for ΔN2-M761 and M761 constructs. The estimated KAD values, obtained from fitting the data to the equation kobs = k0/(1 + [ADP]/KAD), were 15 μm for ΔN2-M761 and 182 μm for M761 (Table 3).
      Functional Analysis—The motor activity of the purified myosins was analyzed using in vitro motility assays (
      • Kron S.J.
      • Spudich J.A.
      ,
      • Kron S.J.
      • Toyoshima Y.Y.
      • Uyeda T.Q.
      • Spudich J.A.
      ). ΔN1-myosin and ΔN2-myosin supported the movement of actin filaments with velocities of 48 and 340 nm/s, respectively. In comparison, actin filaments moved with a velocity of 2.6 μm/s on surfaces decorated with wild-type myosin (Fig. 4).
      Figure thumbnail gr7
      FIGURE 4Sliding velocity of filamentous actin on myosin-coated surfaces. Motor function was analyzed directly by measuring the sliding velocity of rhodamine phalloidin-labeled actin filaments on myosin-coated surfaces under a fluorescence microscope (
      • Anson M.
      • Geeves M.A.
      • Kurzawa S.E.
      • Manstein D.J.
      ). For each experiment, 0.25 mg/ml myosin, 20 nm rhodamine phalloidin-labeled actin, and 2 mm ATP were used. The movement of at least 50 filaments was followed, and the average sliding velocity was determined by analysis of the Gaussian distribution. ΔN-1 myosin moved actin filaments at 48.0 ± 8.5 nm/s; ΔN2-myosin moved actin filaments at 342.7 ± 43.0 nm/s; wild-type myosin moved actin filaments at 2.61 ± 0.3 μm/s at 30 °C.
      Complementation assays with myosin-2 null cells were used to test the ability of the constructs to rescue myosin-2-dependent processes. ΔN1-myosin and ΔN2-myosin were produced at a level similar to that of endogenous myosin-2 in wild-type cells. Dictyostelium myosin-2 null cells display characteristic phenotypic alterations that affect cytokinesis and the multicellular stages of Dictyostelium development. Complementation of myosin null cells with ΔN1-myosin did not rescue any of the myosin-2-specific defects. Transformants were unable to undergo normal cytokinesis or to grow in suspension culture, and their development was blocked at the mound stage (Fig. 5). In contrast, myosin-2 null cells transformed with ΔN2-myosin constructs were phenotypically almost normal. They grew at the same rate as null cells that were transformed with a myosin-2 wild-type construct, and they produced viable spores, although their fruiting bodies were markedly smaller than those formed by Dictyostelium cells producing wild-type myosin-2 (Fig. 5).
      Figure thumbnail gr8
      FIGURE 5Functional characterization of ΔN1-myosin and ΔN2-myosin. A, development cycle of Dictyostelium cells. Fruiting body formation is a myosin-2-dependent process that occurs under starvation conditions. AX3-ORF+ wild-type cells form fruiting bodies that consist of a long stalk and an unwalled, globose spore mass on the top. B, HS1 cells in which the mhcA gene has been deleted are not able to undergo the complete process of fruiting body formation and sporogenesis; instead the cells stop differentiation after they have formed an aggregation mount. C, complementation of myosin null cells with the wild-type mhcA restores the developmental defect. D, null cells producing ΔN1-myosin are not able to form fruiting bodies with viable spores. Their development cycle progresses only to the formation of small stem-like structures without formation of viable spores. E, complementation of myosin null cells with ΔN2-myosin leads to the formation of fruiting bodies with short stalks, normal spore heads, and viable spores. F, cell growth in suspension culture. Individual Dictyostelium cell lines expressing wild-type and mutant myosins were transferred to 100 ml of HL-5C medium at densities of 105 cells/ml and grown on a rotary shaker at 180 rpm. AX3-ORF+ cells (▪) show a doubling time of ∼8 h, whereas myosin null cells (HS1; ○) are unable to grow in suspension. This defect is rescued by the reintroduction of the mhcA gene into the null cells as shown by HS1 cells expressing wild type myosin-2 (□). B, HS1 cells expressing ΔN1-myosin (▾) are not able to rescue the defect, whereas HS1 cells producing ΔN2-myosin (▴) grow at a similar rate as HS1 cells that produce wild-type myosin-2 (□).

      DISCUSSION

      The N-terminal subdomain of myosin-2 is one of the least conserved regions within the myosin head. Although class 1 myosins completely lack this region, it is clear from our in vivo functional analysis that the region or part of it is important for the normal functioning of a class 2 myosin. Kinetic analysis of ΔN1- and ΔN2-myosin constructs revealed that the N-terminal truncations affect but do not abolish the ability of myosin to bind nucleotides, to hydrolyze ATP, or to interact with F-actin. The truncations neither changed the apparent affinity for ATP in the absence of actin nor did they greatly perturb the basal Mg2+-ATPase rate. Actin activation of ATPase activity is also maintained in these constructs.
      The transient kinetic characterization of ΔN1-M761, which lacks the complete N-terminal subdomain, proved to be difficult because of the weakness or complete absence of spectroscopic signals associated with the binding of nucleotides. ΔN1-M761 binds at least 80-fold more weakly to F-actin and dissociates 12 times faster from F-actin in the absence of nucleotides. The weak interaction with F-actin, which is also apparent in the presence of nucleotides, correlates well with a reduced activation of ATPase activity and a reduction in the maximum turnover rate, kcat. In addition, the 40-fold decrease in the rate of mantADP release indicates that ADP release may become the rate-limiting step for basal ATPase, which is also apparent from the close similarity of the rate constants k-D and kbasal measured for ΔN1-M761.
      The head domain construct ΔN2-M761 shows almost wild type-like kinetic properties with binding and release rates for mantADP similar to M761. The observed 6- and 8-fold reductions in catalytic (kcat) and motor activities (Vmax) of ΔN2-myosin can be linked to perturbations in F-actin binding and ADP binding and coupling between the actin and nucleotide binding sites. These perturbations include an 18-fold decrease in F-actin affinity in the absence of nucleotides (KA), a 12-fold increase in affinity for ADP in the actin-bound state (KAD), and an almost 7-fold reduction in the coupling ratio (KAD/KD).
      ΔN1-myosin and ΔN2-myosin support actin filament movement. The 54- and 8-fold reduced velocities correlate well with the observed changes in the rates of ADP release that were observed for ΔN1-myosin and ΔN2-myosin, respectively. In the presence of a saturating number of myosin motors, velocity is proportional to d/ts, where d is the stroke size, and ts is the strongly bound state time. ts is independent of total ATP hydrolysis time and is determined by the rate of ADP release. The reduced velocities can therefore be attributed directly to a slow ADP release rate and a high ADP affinity for actomyosin (
      • Batra R.
      • Geeves M.A.
      • Manstein D.J.
      ).
      The results of in vitro motility assays with ΔN1-myosin and ΔN2-myosin explain at least in part the phenotypic changes observed for myosin null cells producing the truncated myosins. ΔN1-myosin displays a more than 50-fold reduced motility in the assay and is not able to compensate the myosin-dependent defects when produced in null cells. In contrast, ΔN2-myosin displays only a 7-fold reduced motile activity, and complementation assays with ΔN2-myosin indicate almost normal myosin-2 function. This observation suggests that the SH3-like subdomain is not critical for the in vivo function of myosin-2 under the conditions examined.
      Several earlier studies had addressed the interactions of the N-terminal region with other regions of the myosin head fragment. Using changes in intrinsic protein fluorescence, Berger and colleagues (
      • van Duffelen M.
      • Chrin L.R.
      • Berger C.L.
      ,
      • van Duffelen M.
      • Chrin L.R.
      • Berger C.L.
      ) concluded that the SH3-like subdomain of smooth muscle myosin is conformationally sensitive to nucleotide binding and/or hydrolysis. They further suggested that the N-terminal region is only indirectly coupled to the active site but that it is sensitive to direct interactions with the lever arm in the strongly bound states of the ATPase cycle (
      • van Duffelen M.
      • Chrin L.R.
      • Berger C.L.
      ,
      • van Duffelen M.
      • Chrin L.R.
      • Berger C.L.
      ). Contacts between the N terminus and heavy chain residues 750-760 in the converter region (D. discoideum numbering) have been implicated by differential scanning calorimetry as important for the structural integrity and stability of the entire motor domain (
      • Nikolaeva O.P.
      • Orlov V.N.
      • Bobkov A.A.
      • Levitsky D.I.
      ). Following tryptic cleavage of the skeletal muscle myosin heavy chain between Arg23 and Ile24, Levitsky and co-workers (
      • Nikolaeva O.P.
      • Orlov V.N.
      • Bobkov A.A.
      • Levitsky D.I.
      ) observed a reduction of the thermal transition from 49 to 42 °C, whereas nucleotide binding was unaffected. In contrast, tryptic cleavage in loops 1 and 2 affected neither the thermal transition temperature nor nucleotide binding. In agreement with these observations, we found that the ΔN1 constructs displayed an almost complete loss of enzymatic activity within 48 h after their purification. ΔN1-M761 displayed complete loss of activity following freeze-thawing. The importance of contacts between N-terminal residues and residues 750-760 in the converter was supported by similarities in the kinetic behavior of ΔN1-M761 and M754, a motor domain construct that is truncated at position 754 of the myosin heavy chain.
      In conclusion, our experiments show that the initial 33 N-terminal residues are of great importance for normal communication between the functional regions within the myosin-2 motor domain. Based on the observed differences between the crystal structures of class 1 and class 2 motor domains (
      • Kollmar M.
      • Durrwang U.
      • Kliche W.
      • Manstein D.J.
      • Kull F.J.
      ,
      • Fisher A.J.
      • Smith C.A.
      • Thoden J.B.
      • Smith R.
      • Sutoh K.
      • Holden H.M.
      • Rayment I.
      ,
      • Reubold T.F.
      • Eschenburg S.
      • Becker A.
      • Kull F.J.
      • Manstein D.J.
      ), we suggest that the N-terminal residues Pro3, Ile4, Tyr11, Leu15, and Tyr14 play an important role because they stabilize the motor domain through clustered hydrophobic interactions with residues Pro133, Ile134, Met139, Ile142, Phe143, and His154 in the proximal part of the N-terminal region (Fig. 1, B and C). Additional structural stability in this region is maintained by hydrophobic interactions formed between residues Phe25-Lys32 and residues 760-765 on an adjacent helix that emerges from the converter region (Fig. 1A). Consequently, noncovalent attractive forces between the N-terminal and C-terminal regions of the motor domain are important for the structural and functional integrity of the myosin motor. Removal of only the SH3-like subdomain, as performed in the ΔN-2 constructs, does not affect the biological function of myosin-2 in Dictyostelium. However, the replacement of the SH3-like subdomain still has a clear effect on the way in which conformational changes that follow the binding of F-actin or nucleotide are communicated within the motor domain. At this stage, it is tempting to speculate that the mechanical perturbations of the SH3-like subdomain, as they may occur in the sarcomere, in the crowded context of a cleavage furrow, or in other physiological situations, can lead to changes in the functional behavior of the motor that are similar in extent or even greater than those observed for ΔN-2-myosin. The way in which intra- and intermolecular contacts (
      • van Duffelen M.
      • Chrin L.R.
      • Berger C.L.
      ) and modifications and substitutions of light chains (
      • Terrak M.
      • Rebowski G.
      • Lu R.C.
      • Grabarek Z.
      • Dominguez R.
      ) may affect or modulate the communication pathway via the SH3-like subdomain and whether such a pathway is important for the regulation and functional fine-tuning of at least some myosins remain to be elucidated in further studies.

      Acknowledgments

      We thank R. Schumann and S. Zimmermann for excellent technical assistance, R. Fedorov for providing Fig. 1, and M. A. Geeves, N. Tzvetkov, and R. Fedorov for discussions.

      References

        • Sellers J.R.
        Biochim. Biophys. Acta. 2000; 1496: 3-22
        • Holmes K.C.
        • Schröder R.R.
        • Sweeney H.L.
        • Houdusse A.
        Philos. Trans. R. Soc. Lond. B Biol. Sci. 2004; 359: 1819-1828
        • Geeves M.A.
        • Fedorov R.
        • Manstein D.J.
        Cell. Mol. Life Sci. 2005; 62: 1462-1477
        • Menetrey J.
        • Bahloul A.
        • Wells A.L.
        • Yengo C.M.
        • Morris C.A.
        • Sweeney H.L.
        • Houdusse A.
        Nature. 2005; 435: 779-785
        • Coureux P.D.
        • Wells A.L.
        • Menetrey J.
        • Yengo C.M.
        • Morris C.A.
        • Sweeney H.L.
        • Houdusse A.
        Nature. 2003; 425: 419-423
        • Smith C.A.
        • Rayment I.
        Biochemistry. 1996; 35: 5404-5417
        • Schroder R.R.
        • Manstein D.J.
        • Jahn W.
        • Holden H.
        • Rayment I.
        • Holmes K.C.
        • Spudich J.A.
        Nature. 1993; 364: 171-174
        • Dominguez R.
        • Freyzon Y.
        • Trybus K.M.
        • Cohen C.
        Cell. 1998; 94: 559-571
        • Kollmar M.
        • Durrwang U.
        • Kliche W.
        • Manstein D.J.
        • Kull F.J.
        EMBO J. 2002; 21: 2517-2525
        • Koch C.A.
        • Anderson D.
        • Moran M.F.
        • Ellis C.
        • Pawson T.
        Science. 1991; 252: 668-674
        • Onishi H.
        • Maita T.
        • Matsuda G.
        • Fujiwara K.
        J. Biol. Chem. 1990; 265: 19362-19368
        • Ito K.
        • Liu X.
        • Katayama E.
        • Uyeda T.Q.
        Biophy. J. 1999; 76: 985-992
        • Terrak M.
        • Rebowski G.
        • Lu R.C.
        • Grabarek Z.
        • Dominguez R.
        Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 12718-12723
        • Manstein D.J.
        • Schuster H.P.
        • Morandini P.
        • Hunt D.M.
        Gene. 1995; 162: 129-134
        • Manstein D.J.
        • Hunt D.M.
        J. Muscle Res. Cell Motil. 1995; 16: 325-332
        • Kurzawa S.E.
        • Manstein D.J.
        • Geeves M.A.
        Biochemistry. 1997; 36: 317-323
        • Manstein D.J.
        • Titus M.A.
        • De Lozanne A.
        • Spudich J.A.
        EMBO J. 1989; 8: 923-932
        • Egelhoff T.T.
        • Titus M.A.
        • Manstein D.J.
        • Ruppel K.M.
        • Spudich J.A.
        Methods Enzymol. 1991; 196: 319-334
        • Ruppel K.M.
        • Uyeda T.Q.
        • Spudich J.A.
        J. Biol. Chem. 1994; 269: 18773-18780
        • Criddle A.H.
        • Geeves M.A.
        • Jeffries T.
        Biochem. J. 1985; 232: 343-349
        • Anson M.
        • Geeves M.A.
        • Kurzawa S.E.
        • Manstein D.J.
        EMBO J. 1996; 15: 6069-6074
        • Uyeda T.Q.
        • Kron S.J.
        • Spudich J.A.
        J. Mol. Biol. 1990; 214: 699-710
        • Batra R.
        • Geeves M.A.
        • Manstein D.J.
        Biochemistry. 1999; 38: 6126-6134
        • Furch M.
        • Fujita-Becker S.
        • Geeves M.A.
        • Holmes K.C.
        • Manstein D.J.
        J. Mol. Biol. 1999; 290: 797-809
        • Tsiavaliaris G.
        • Fujita-Becker S.
        • Batra R.
        • Levitsky D.I.
        • Kull F.J.
        • Geeves M.A.
        • Manstein D.J.
        EMBO Rep. 2002; 3: 1099-1105
        • Furch M.
        • Geeves M.A.
        • Manstein D.J.
        Biochemistry. 1998; 37: 6317-6326
        • Woodward S.K.
        • Geeves M.A.
        • Manstein D.J.
        Biochemistry. 1995; 34: 16056-16064
        • Kron S.J.
        • Spudich J.A.
        Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 6272-6276
        • Kron S.J.
        • Toyoshima Y.Y.
        • Uyeda T.Q.
        • Spudich J.A.
        Methods Enzymol. 1991; 196: 399-416
        • van Duffelen M.
        • Chrin L.R.
        • Berger C.L.
        Biochem. Biophys. Res. Commun. 2005; 329: 563-572
        • van Duffelen M.
        • Chrin L.R.
        • Berger C.L.
        Biophys. J. 2004; 87: 1767-1775
        • Nikolaeva O.P.
        • Orlov V.N.
        • Bobkov A.A.
        • Levitsky D.I.
        Eur. J. Biochem. 2002; 269: 5678-5688
        • Fisher A.J.
        • Smith C.A.
        • Thoden J.B.
        • Smith R.
        • Sutoh K.
        • Holden H.M.
        • Rayment I.
        Biochemistry. 1995; 34: 8960-8972
        • Reubold T.F.
        • Eschenburg S.
        • Becker A.
        • Kull F.J.
        • Manstein D.J.
        Nat. Struct. Biol. 2003; 10: 826-830