Portal de Periódicos da CAPES

The Myosin Relay Helix to Converter Interface Remains Intact throughout the Actomyosin ATPase Cycle

2001; Elsevier BV; Volume: 276; Issue: 22 Linguagem: Inglês

10.1074/jbc.m010887200

1083-351X

William M. Shih, James A. Spudich,

Cardiovascular Effects of Exercise

Crystal structures of the myosin motor domain in the presence of different nucleotides show the lever arm domain in two basic angular states, postulated to represent prestroke and poststroke states, respectively (Rayment, I. (1996) J. Biol. Chem. 271, 15850–15853; Dominguez, R., Freyzon, Y., Trybus, K. M., and Cohen, C. (1998) Cell 94, 559–571). Contact is maintained between two domains, the relay and the converter, in both of these angular states. Therefore it has been proposed by Dominguez et al. (cited above) that this contact is critical for mechanically driving the angular change of the lever arm domain. However, structural information is lacking on whether this contact is maintained throughout the actin-activated myosin ATPase cycle. To test the functional importance of this interdomain contact, we introduced cysteines into the sequence of a "cysteine-light" myosin motor at position 499 on the lower cleft and position 738 on the converter domain (Shih, W. M., Gryczynski, Z., Lakowicz, J. L., and Spudich, J. A. (2000) Cell 102, 683–694). Disulfide cross-linking could be induced. The cross-link had minimal effects on actin binding, ATP-induced actin release, and actin-activated ATPase. These results demonstrate that the relay/converter interface remains intact in the actin strongly bound state of myosin and throughout the entire actin-activated myosin ATPase cycle. Crystal structures of the myosin motor domain in the presence of different nucleotides show the lever arm domain in two basic angular states, postulated to represent prestroke and poststroke states, respectively (Rayment, I. (1996) J. Biol. Chem. 271, 15850–15853; Dominguez, R., Freyzon, Y., Trybus, K. M., and Cohen, C. (1998) Cell 94, 559–571). Contact is maintained between two domains, the relay and the converter, in both of these angular states. Therefore it has been proposed by Dominguez et al. (cited above) that this contact is critical for mechanically driving the angular change of the lever arm domain. However, structural information is lacking on whether this contact is maintained throughout the actin-activated myosin ATPase cycle. To test the functional importance of this interdomain contact, we introduced cysteines into the sequence of a "cysteine-light" myosin motor at position 499 on the lower cleft and position 738 on the converter domain (Shih, W. M., Gryczynski, Z., Lakowicz, J. L., and Spudich, J. A. (2000) Cell 102, 683–694). Disulfide cross-linking could be induced. The cross-link had minimal effects on actin binding, ATP-induced actin release, and actin-activated ATPase. These results demonstrate that the relay/converter interface remains intact in the actin strongly bound state of myosin and throughout the entire actin-activated myosin ATPase cycle. cysteine-light polyacrylamide gel electrophoresis Myosins are molecular motors that transduce the chemical energy of ATP hydrolysis to mechanical work in the form of the vectorial translocation of substrate actin filaments. In the swinging lever arm model of actin-myosin motor action, myosin binds to the actin with its globular catalytic domain and then rotates its carboxyl-terminal lever arm domain (reviewed in Ref. 5Goldman Y.E. Cell. 1998; 93: 1-4Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar). The anchoring of the end of the lever arm domain results in the translocation of the catalytic domain and the attached actin filament. Crystal structures of the motor domain of myosin show the lever arm domain in two basic angular classes, which have been postulated to represent prestroke and poststroke states (2Smith C.A. Rayment I. Biochemistry. 1996; 35: 5404-5417Crossref PubMed Scopus (515) Google Scholar, 3Dominguez R. Freyzon Y. Trybus K.M. Cohen C. Cell. 1998; 94: 559-571Abstract Full Text Full Text PDF PubMed Scopus (591) Google Scholar, 6Rayment 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). Scallop myosin complexed with ADP has been crystallized in a third angular state, proposed as a myosin·ATP actin-detached state (7Houdusse A. Kalabokis V.N. Himmel D. Szent-Gyorgyi A.G. Cohen C. Cell. 1999; 97: 459-470Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar). Recent dynamic studies of the lever arm position using steady state and time-resolved fluorescence energy transfer measurements support a swing of the lever arm from a prestroke state to a poststroke state through an angle of more than 70 degrees (4Shih W.M. Gryczynski Z. Lakowicz J.L. Spudich J.A. Cell. 2000; 102: 683-694Abstract Full Text Full Text PDF PubMed Scopus (125) Google Scholar). The lever arm domain consists of a disc-shaped "converter domain" from which a long α-helix, bound by two calmodulin-like light chains, emerges. In both crystal states, the converter domain maintains a contact on a face of its radial edge to a rigid helix extending from the lower domain of the large cleft, referred to as the relay helix (3Dominguez R. Freyzon Y. Trybus K.M. Cohen C. Cell. 1998; 94: 559-571Abstract Full Text Full Text PDF PubMed Scopus (591) Google Scholar). This helix has to undergo a small conformational change, primarily a rigid body translation and rotation, to accommodate the angular rotation of the converter domain. Because this domain-domain interface is maintained between the two angular states, it has been proposed that this interface is important for mechanically driving the angular change. However, structural information is lacking as to whether this contact is maintained throughout the actin-activated myosin ATPase cycle, including in an actin strong binding state. We have used a cysteine engineering approach to address the question as to whether the relay/converter contact is maintained in the actin strongly bound state. We constructed myosinII alleles containing cysteine-light mutations (C49S, C312Y, C442S, C470I, C599L, and C678Y) (4Shih W.M. Gryczynski Z. Lakowicz J.L. Spudich J.A. Cell. 2000; 102: 683-694Abstract Full Text Full Text PDF PubMed Scopus (125) Google Scholar) and substituted cysteine codons into positions corresponding to either residue 499 or 738 or both (Fig. 1). We show that the mutant myosins are functional in vivo and in vitro. The cross-linked myosin (containing cysteines at both positions 499 and 738) retained the ability to bind to actin in the absence of ATP as well as the ability to be released from actin in the presence of ATP. The cross-linked myosin also retained actin-activated ATPase activity. The changes made in a given mutant myosin are described within parentheses. For example, myosin(CL, I499C) refers to a full-length myosin gene with the cysteine-light (CL)1 mutations (C49S, C312Y, C442S, C470I, C599L, and C678Y) (4Shih W.M. Gryczynski Z. Lakowicz J.L. Spudich J.A. Cell. 2000; 102: 683-694Abstract Full Text Full Text PDF PubMed Scopus (125) Google Scholar) and the mutation I499C. Subcloning procedures were carried out using standard protocols (8Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory Press, New York1989Google Scholar). myosin(CL, I499C), myosin(CL, R738C), and myosin(CL, I499C, R738C) were generated by splice overlap extension mutagenesis (9Ho S.N. Hunt H.D. Horton R.M. Pullen J.K. Pease L.R. Gene (Amst.). 1989; 77: 51-59Crossref PubMed Scopus (6851) Google Scholar) using myosin(CL) as a template. Myosin genes were subcloned into the expression vector pTIKL-Myo. The introduced mutations were verified by dideoxy-DNA sequencing. The S1 gene fragments were then subcloned into pTIKLOES1, an expression vector for producing S1 with a carboxyl-terminal His6 tag on the heavy chain. The following oligonucleotides were used for mutagenesis: I499C-F, 5′-TATCTTAAAGAGAAATGTAATTGGACTTTCATC-3′; I499C-R, 5′-GATGAAAGTCCAATTACATTTCTCTTTAAGATA-3′; R738C-F, 5′-GATCCAGAACAATATTGTTTCGGTATCACCAAG-3′; and R738C-R, 5′-CTTGGTGATACCGAAACAATATTGTTCTGGATC-3′. Dictyostelium cells were grown in HL-5 medium as described previously (10Murphy C.T. Spudich J.A. Biochemistry. 1998; 37: 6738-6744Crossref PubMed Scopus (83) Google Scholar). Cells were grown at 22 °C in HL-5 supplemented with 17% FM medium (Life Technologies, Inc.), 100 units/ml penicillin, and 100 units/ml streptomycin. Transformations were performed as described previously (11Egelhoff T.T. Titus M.A. Manstein D.J. Ruppel K.M. Spudich J.A. Methods Enzymol. 1991; 196: 319-334Crossref PubMed Scopus (72) Google Scholar). The mhcA null cell line HS1 was transformed with 10 μg of each of the pTIKL-Myo plasmids bearing wild type or mutant versions of the full-length myosin, whereas the AX3-ORF+ cell line was transformed with 10 μg of each of the pTIKLOES1 plasmids bearing wild type or mutant versions of the S1 fragment of myosin. Clonal cell lines that grew in the presence of HL-5 supplemented with penicillin, streptomycin, and 8 μg/ml G418 were isolated, and these cell lines were further characterized. Cells were grown on plates to near confluence before they were transferred to shaking flasks. Cells were diluted to 4 × 104 cells/ml in 25 ml of total volume HL-5 in 125-ml Erlenmeyer flasks and shaken at 200 rpm at 22 °C for 6 days. A small aliquot was removed at regular intervals, and the number of cells was counted using a hemocytometer. Dictyostelium S1 His6 was expressed in DictyosteliumAX3-ORF+ cells (grown in suspension) and purified as described (12Giese K.C. Spudich J.A. Biochemistry. 1997; 36: 8465-8473Crossref PubMed Scopus (33) Google Scholar). S1 ATPase activities were measured as the release of labeled Pi using γ-32P, as reported previously (12Giese K.C. Spudich J.A. Biochemistry. 1997; 36: 8465-8473Crossref PubMed Scopus (33) Google Scholar). The plotted points and error bars of Fig. 5 represent measurements from three independent trials for each of two different protein preparations. Cross-linking was induced by the addition of 25 μm 5,5′-dithiobis(nitrobenzoic acid) to S1 at a concentration of 1–2 μm. The buffer conditions were 25 mm HEPES, pH 7.0, 25 mmNaCl, and 10 mm MgCl2. The cross-linking reaction was quenched by the addition of 1 mmdithiothreitol. (The Cys499-Cys738 disulfide cross-link is not reduced by 1 mm dithiothreitol under nondenaturing conditions.) Disulfide cross-linking was assayed by mobility shift in SDS-PAGE behavior in the absence of reducing agent. Confirmation of disulfide bond formation was made by SDS-PAGE analysis in the presence of reducing agent reversing the mobility shift. The buffer conditions used were 25 mm HEPES, pH 7.0, 25 mm NaCl, and 10 mm MgCl2. For the noncompetitive assay, S1 (0.8 μm final concentration) was mixed with F-actin (3.0 μm final concentration) for 10 min, and then the mixture was centrifuged at 100,000 × g for 10 min. The supernatant and the resuspended pellet were examined by SDS-PAGE to determine whether the S1 cosedimented with the F-actin. The same assay was repeated in the presence of 2 mm Mg-ATP. For the competitive assay, the same procedure was used but with S1 at a final concentration of 1.3 μm and F-actin at a concentration of 0.5 μm. Dictyostelium cells that lack the myosinII gene are unable to divide in suspension, instead becoming large and multinucleate before eventually lysing and dying (13De Lozanne A. Spudich J.A. Science. 1987; 236: 1086-1091Crossref PubMed Scopus (763) Google Scholar, 14Knecht D.A. Loomis W.F. Science. 1987; 236: 1081-1086Crossref PubMed Scopus (507) Google Scholar). Thus the transformation of the mutant myosinII gene into these myosinII null cells can lead to a simple assay for in vivo function (assaying for the rescue of the growth in suspension defect). The ability to rescue an in vivo defect serves as a useful benchmark to demonstrate that mutant myosins behave in a functional manner. The design of this experiment relies on the un-cross-linked double-cysteine mutant myosin to behave in a functional manner and then to assay for biochemical differences upon specific cross-linking. Arg738 (with its β and γ carbons) and Ile499 form part of the hydrophobic interface between the relay helix and the converter domain, respectively (Fig.1). Disruption of this hydrophobic interface thus could potentially be destabilizing for the myosin. Cysteines are relatively hydrophobic, however, and should be good candidates for replacement side chains for packing in the interface. Fig. 2 shows a growth curve examining the growth of Dictyostelium cells that were missing the genomic copy of myosinII but were supplied with another copy on an extrachromosomal plasmid. All of the mutant myosinII genes introduced (myosin(CL), myosin(CL, I499C), myosin(CL, R738C), and myosin(CL, I499C, R738C)) rescued growth in suspension to a rate comparable with that of the wild type. The parent strain lacking a copy of myosinII, however, failed to grow in suspension. Therefore it appears that the introduction of either cysteine, both individually or in tandem, is well tolerated by the structure of myosin. According to the crystal structures available, the side chains of residues at positions 499 and 738 are in close proximity (2Smith C.A. Rayment I. Biochemistry. 1996; 35: 5404-5417Crossref PubMed Scopus (515) Google Scholar, 15Fisher A.J. Smith C.A. Thoden J.B. Smith R. Sutoh K. Holden H.M. Rayment I. Biochemistry. 1995; 34: 8960-8972Crossref PubMed Scopus (637) Google Scholar). Previous studies have shown that cysteines placed at nearby positions in a structure usually can be induced to form a disulfide cross-link, which is catalyzed either by ambient oxygen or by a disulfide exchange reagent (16Falke J.J. Koshland D.E.J. Science. 1987; 237: 1596-1600Crossref PubMed Scopus (228) Google Scholar). The cross-linking of two residues in the structure of a protein that are separated by a large number of residues results in a covalently closed large loop within the primary sequence. Denatured proteins containing such a loop might be expected to exhibit a different mobility during gel electrophoresis; examples have been found where cross-linking induces either a gel mobility increase (17Careaga C.L. Falke J.F. J. Mol. Biol. 1992; 226: 1219-1235Crossref PubMed Scopus (218) Google Scholar) or a gel mobility decrease (18Murai N. Makino Y. Yoshida M. J. Biol. Chem. 1996; 271: 28234-28829Abstract Full Text Full Text PDF Scopus (47) Google Scholar). Fig. 3 shows that treatment with the disulfide exchange reagent dithionitrobenzoate induces an SDS-PAGE mobility increase in over 85% of a cysteine-light myosin S1 only when cysteines are introduced at both positions 499 and 738 but not when the cysteines are introduced individually. This gel mobility shift is reversible if the protein is loaded onto the gel in the presence of a reducing agent such as β-mercaptoethanol. Thus a specific cross-link is formed between introduced cysteines at positions 499 and 738. Curiously, the disulfide cross-link that is formed is quite robust against reduction under native conditions. Overnight treatment in the presence of 1 mm dithiothreitol, 1 mmβ-mercaptoethanol, or 1 mm tricarboxyethylphosphine followed by passage through a Sephadex spin column (to remove the dithiothreitol) fails to significantly reduce the disulfide, as analyzed by SDS-PAGE (data not shown). Only after SDS denaturation does the disulfide become accessible to reduction. In the absence of the disulfide exchange reagent and reducing agents, the un-cross-linked protein is converted slowly to the cross-linked form over a period of weeks when stored at 4 °C, presumably catalyzed by ambient dissolved oxygen (data not shown). Wild type S1 precipitates with F-actin in the absence of ATP during centrifugation at sufficient speeds. Small, unattached proteins sediment much more slowly and therefore remain in the supernatant. This assay can be used to investigate the binding of S1 to actin. The property of the cross-linked form of S1 (that it migrates more rapidly during SDS-PAGE, relative to the un-cross-linked form) lends itself as a special advantage in the F-actin cosedimentation assay because this assay allows for the analysis of the supernatant (actin-detached) and pellet (actin-attached) fractions using SDS-PAGE. Thus the behavior of the two forms can be analyzed simultaneously in the same sample volume. Fig. 4 A shows S1 binding to F-actin in the presence of excess F-actin and in the absence of ATP. Both the cross-linked and un-cross-linked forms of S1 cosediment with the F-actin. Fig. 4 A also shows S1 binding to F-actin in the presence of excess F-actin and ATP. Both the cross-linked and un-cross-linked forms of S1 are similarly released from the actin by the ATP. Fig. 4 B shows S1 binding to F-actin in the presence of limiting F-actin. If the cross-linked form of S1 had a significantly different binding affinity to F-actin as compared with the un-cross-linked form of S1, then the ratio of the un-cross-linked to cross-linked bands should be significantly different in the supernatant and pellet lanes. However, Fig. 4 B shows that no significant difference is evident. Thus the cross-link does not perturb the binding of myosin to F-actin in its strong affinity mode. The rate-limiting step in the ATPase cycle of myosin in the absence of F-actin is the product release (19Lymn R.W. Taylor E.W. Biochemistry. 1970; 9: 2975-2983Crossref PubMed Scopus (200) Google Scholar). The presence of F-actin stimulates the myosin to release its products more rapidly (20Lymn R.W. Taylor E.W. Biochemistry. 1971; 10: 4617-4624Crossref PubMed Scopus (1021) Google Scholar). We examined the actin-activated ATPase activity of the cross-linked form versus the un-cross-linked form to see if the actin stimulation of product release might be impaired. Fig. 5 shows that the actin activation of the cross-linked species is decreased by 29% and that the K m for the actin activation is decreased by 29%. The basal rate of ATP hydrolysis (i.e. in the absence of F-actin) is the same in the cross-linked and un-cross-linked forms. Thus phosphate release in the cross-linked myosin is still stimulated by F-actin, although the amount of the stimulation is slightly decreased. The crystal structures of the Dictyostelium myosinII catalytic domain have been solved in complex with a number of nucleotide analogues, including ADP beryllium fluoride (BeFx), ADP aluminum fluoride (AlF4−), and ADP vanadate (Vi), all thought to mimic myosin states with a weak affinity for actin. It is only upon phosphate release that the myosin can adopt a conformation that allows it to bind to actin with a strong affinity. The crystal structures fall into two basic conformational classes: one that is thought to represent the myosin with its lever arm in a putative poststroke angle and one that is thought to represent the myosin with its lever arm in a putative prestroke angle. The lever arm extends off from a 60-residue globular domain, often referred to as the converter domain. This converter domain shares in both structures a hydrophobic interface with the back end of the lower jaw of the large cleft of the catalytic domain, referred to as the relay helix, despite a 70° rotation of the converter domain between the two structures. Fig. 6 shows this interface in the two structures, as well as in an intermediate structure that was generated as an interpolation between the crystal structure models (4Shih W.M. Gryczynski Z. Lakowicz J.L. Spudich J.A. Cell. 2000; 102: 683-694Abstract Full Text Full Text PDF PubMed Scopus (125) Google Scholar). Inspection of the crystal structures suggests a mechanical picture describing the sequence of forces that are behind the power stroke. It has been proposed that changes in the nucleotide binding site, triggered by the formation of a bond between the backbone amide of switch II residue Gly457 with the γ-phosphate in the prestroke state, are amplified into a larger change in the position of the relay helix. This change drives a rotation and translation of the lever arm through its converter domain (3Dominguez R. Freyzon Y. Trybus K.M. Cohen C. Cell. 1998; 94: 559-571Abstract Full Text Full Text PDF PubMed Scopus (591) Google Scholar, 15Fisher A.J. Smith C.A. Thoden J.B. Smith R. Sutoh K. Holden H.M. Rayment I. Biochemistry. 1995; 34: 8960-8972Crossref PubMed Scopus (637) Google Scholar). These structures, however, are thought to represent myosin in conformations that have a weak affinity to actin (21Werber M.M. Peyser Y.M. Muhlrad A. Biochemistry. 1992; 31: 7190-7197Crossref PubMed Scopus (125) Google Scholar). Myosin undergoes a conformational change that allows it to bind to actin with a much higher affinity; this conformational change in myosin can be monitored by the pyrene labeling of Cys374 on actin (22Criddle A.H. Geeves M.A. Jeffries T. Biochem. J. 1985; 232: 343-349Crossref PubMed Scopus (180) Google Scholar) or with mant-ADP in the active site of the myosin (23Woodward S.K. Eccleston J.F. Geeves M.A. Biochemistry. 1991; 30: 422-430Crossref PubMed Scopus (135) Google Scholar). These spectroscopic probes, however, give no structural details of the nature of this conformational change. Therefore currently there is no experimental evidence that allows one to build a reliable model of myosin in its strong affinity to actin state. One proposal has been that the large cleft completely closes to achieve this strong affinity to actin state (1Rayment I. J. Biol. Chem. 1996; 271: 15850-15853Abstract Full Text Full Text PDF PubMed Scopus (98) Google Scholar, 6Rayment 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). One possible extension of this model is that the complete closure of the large cleft causes the relay helix to slip away from the converter domain. The converter domain, after its release from the relay helix, snaps back to its poststroke conformation, thus completing the power stroke. In this model, the lever arm behaves as a torque spring that is wound by 70° by the back end of the lower jaw and then suddenly released once actin strong binding has been achieved. Another model is that the interface between the converter domain and the relay helix is maintained in the strong actin affinity state of myosin as well. In this case, the relay helix may mechanically transmit the changes to the lever arm both in the actin-detached stages (the recovery stroke) and in the actin-attached stages (the power stroke). If the first model were correct (that the interface must slip), then the cross-linking of the interface would prevent that slipping and should prevent the myosin from achieving its actin strong binding affinity state. On the other hand, if the second model were correct (that the interface is maintained during the whole cycle), then the cross-linking of the interface should have little or no effect on the transition to the strong affinity to actin state. Our experiments demonstrate that cross-linking these two domains together through a cysteine at position 499 on the relay helix and a cysteine at position 738 on the converter domain does not inhibit myosin from achieving a strong affinity to actin state nor does it inhibit the effect of ATP in shifting the myosin back to the weak affinity to actin state. Thus our results are consistent with the model that the interface is maintained both in the actin-detached and actin-attached stages of the actomyosin ATPase cycle. We thank András Málnási-Csizmadia and Clive Bagshaw for an advance copy of their manuscript and discussions. We thank Wen Liang for assistance with generating the growth curves, and we thank Doug Robinson for helpful discussions.