Portal de Periódicos da CAPES

Formation of Stable Inhibitory Complexes of Myosin Subfragment 1 Using Fluoroscandium Anions

1995; Elsevier BV; Volume: 270; Issue: 33 Linguagem: Inglês

10.1074/jbc.270.33.19282

1083-351X

Damodaragounder Gopal, Morris Burke,

Ion channel regulation and function

Evidence is presented that MgADP can be noncovalently trapped in myosin subfragment 1 in the presence of ScFx resulting in the concomitant loss of ATPase function. The rate of inactivation in the presence of MgCl2 at 25°C is 8.7 M−1 s−1 which is too slow for a simple collisional mechanism and suggests that a subsequent slow isomerization step is responsible for formation of a stable ternary complex, S1•MgADP•ScFx in a manner analogous to that proposed for the Vi stabilized complex by 12Goodno C.C. Proc. Natl. Acad. Sci. U. S. A. 1979; 76: 2620-2624Google Scholar Proc. Natl. Acad. Sci. U. S. A. 76, 2620-2624). It is also found that ADP can be trapped in subfragment 1 in the absence of MgCl2 indicating the formation of an S1•ADP•ScFx complex. The stability of these complexes at 4°C was studied by following the loss of trapped [14C]ADP with a chase with ADP. The rate of nucleotide loss at 4°C was biphasic for both complexes suggesting that the inhibitory complexes exist in two distinct states as previously proposed for the ternary complex stabilized by Vi (22Mihashi K. Ooi A. Hiratsuka T. J. Biochem. (Tokyo). 1990; 107: 464-469Google Scholar J. Biochem. (Tokyo) 107, 464-469). Formation of these complexes resulted in a marked enhancement of the intrinsic tryptophyl fluorescence suggesting that conformationally they may resemble the steady-state intermediate formed with MgATP. The failure to observe photolysis in the presence of excess Vi at sites associated with the ATP consensus sequence suggests that in these complexes ScFx occupies the site responsible for these cleavage reactions and that it is not displaced by the added Vi. Evidence is presented that MgADP can be noncovalently trapped in myosin subfragment 1 in the presence of ScFx resulting in the concomitant loss of ATPase function. The rate of inactivation in the presence of MgCl2 at 25°C is 8.7 M−1 s−1 which is too slow for a simple collisional mechanism and suggests that a subsequent slow isomerization step is responsible for formation of a stable ternary complex, S1•MgADP•ScFx in a manner analogous to that proposed for the Vi stabilized complex by 12Goodno C.C. Proc. Natl. Acad. Sci. U. S. A. 1979; 76: 2620-2624Google Scholar Proc. Natl. Acad. Sci. U. S. A. 76, 2620-2624). It is also found that ADP can be trapped in subfragment 1 in the absence of MgCl2 indicating the formation of an S1•ADP•ScFx complex. The stability of these complexes at 4°C was studied by following the loss of trapped [14C]ADP with a chase with ADP. The rate of nucleotide loss at 4°C was biphasic for both complexes suggesting that the inhibitory complexes exist in two distinct states as previously proposed for the ternary complex stabilized by Vi (22Mihashi K. Ooi A. Hiratsuka T. J. Biochem. (Tokyo). 1990; 107: 464-469Google Scholar J. Biochem. (Tokyo) 107, 464-469). Formation of these complexes resulted in a marked enhancement of the intrinsic tryptophyl fluorescence suggesting that conformationally they may resemble the steady-state intermediate formed with MgATP. The failure to observe photolysis in the presence of excess Vi at sites associated with the ATP consensus sequence suggests that in these complexes ScFx occupies the site responsible for these cleavage reactions and that it is not displaced by the added Vi. INTRODUCTIONThe energy for muscle contraction is derived from the hydrolysis of MgATP by myosin and the activation of this process by actin. During this process, myosin forms a series of intermediates resulting in the formation of a steady-state intermediate MADP•Pi which upon reassociation with actin forms AMADP•Pi, a non-force-producing state, which subsequently isomerizes to force-bearing states with and without Pi (AMADP•Pi or AMADP), respectively (16Hibberd M.G. Trentham D.R. Annu. Rev. Biophys. Biophys. Chem. 1986; 15: 119-161Google Scholar; 9Dantzig J.A. Goldman Y.E. Millar N.C. Lacktis J. Homsher E. J. Physiol. 1992; 451: 247-248Google Scholar). Thus, a key question related to the structural basis of force generation is how different is the S1 1The abbreviations used are: S1myosin subfragment 1VivanadateBeFAlF, and ScF are generalized representations of fluorometal complexes formed with beryllium, aluminum, and scandium. structure in the MADP•Pi and MADP•Pi or MADP states. Unfortunately, the former two intermediates have only limited lifetimes, and it appears that the latter state when bound to actin is different in structure from that formed by the binding of MgADP to actoS1 (29Sleep J.A. Hutton R.L. Biochemistry. 1980; 19: 1276-1283Google Scholar). Therefore, stable analogs corresponding to these species would be useful in obtaining information of their structures.Since the demonstration by Goodno(1979) that orthovanadate ion in the presence of MgADP was able to form a stable ATPase-inhibitory complex with myosin S1, corresponding to a 1:1:1 ternary complex designated MADP•Vi, much interest has been devoted in its characterization as it appears to have some properties in common with the steady-state intermediate, MADP•Pi, of the MgATPase hydrolysis cycle by S1 (13Goodno C.C. Methods Enzymol. 1982; 85B: 116-123Google Scholar; 30Smith S.J. Eisenberg E. Eur. J. Biochem. 1990; 193: 69-73Google Scholar; see also 14Goodno C.C. Taylor E.R.W. Proc. Natl. Acad. Sci. U. S. A. 1982; 79: 21-25Google Scholar for a review)). Because of the simple stoichiometry of the complex and because of its inhibitory nature, it is generally assumed that the nucleoside diphosphate resides at the nucleotide binding site with the Vi taking the place of the terminal γ Pi group and, in this way, forms a stable analog of the transition state. Such an interpretation is also supported by the fact that structurally the vanadate ion is similar to orthophosphate (19Lindquist R.N. Lymn J.L. Lienhard G.E. J. Am. Chem. Soc. 1973; 95: 8762-8768Google Scholar).More recently, the use of other complex anions with similarity to Pi has been investigated for their ability to stabilize the transition states of phosphotransferases and nucleotidases in the presence of nucleotide diphosphate (5Chabre M. Trends Biochem. Sci. 1990; 15: 6-10Google Scholar). These include fluoride complexes of beryllium and aluminum, BeFx and AlFx, respectively, and their efficacy in forming stable transition state analog states has been studied with actin (6Combeau C. Carlier M.-F. J. Biol. Chem. 1989; 264: 19017-19021Google Scholar, 1990), with smooth muscle heavy meromyosin (21Maruta S. Henry G.D. Sykes B.D. Ikebe M. J. Biol. Chem. 1993; 268: 7093-7100Google Scholar) and skeletal myosin S1 (26Phan B. Reisler E. Biochemistry. 1992; 31: 4787-4793Google Scholar; 3Beck T.W. Chase P.B. Kushmerick M.J. Biophys. J. 1992; 61: 2506aGoogle Scholar: 35Werber M.M. Peyser Y.M. Muhlrad A. Biochemistry. 1992; 31: 7190-7197Google Scholar; 15Henry G.D. Maruta S. Ikebe M. Sykes B.D. Biochemistry. 1993; 32: 10451-10456Google Scholar).In the present paper we present evidence that fluoroscandium anion (presumably ScFx) can also lead to effective trapping of ADP in S1 in both the presence and absence of Mg2+ with concomitant inactivation of the ATPase properties of the protein. Surprisingly, free nucleoside diphosphate is more stably trapped than the MgADP form with fluoroscandium. Actin binding to these complexes results in a faster release of trapped ADP in accord with the usual effect of actin in accelerating product release from S1.MATERIALS AND METHODSDistilled water further purified through a Millipore QTM system was used throughout. ATP, ADP, dithioerythritol, HEPES, Tris, NaF, KF, sodium orthovanadate, BeCl2, AlCl3, and chymotrypsin were from Sigma and Aldrich. ScCl3 was purchased from Johnson Mathey Electronics with a purity of 99.9%. A stock solution of ScCl3 (100 mM) was made in water and stored at −20°C. [14C]ADP (specific activity 2.4 × 1013 cpm/mol) was purchased from DuPont NEN. All other reagents were of analytical grade.Protein PreparationThe preparations of myosin and S1 were done as described by 10Godfrey J.E. Harrington W.F. Biochemistry. 1970; 9: 886-893Google Scholar and 34Weeds A.G. Taylor R.S. Nature. 1975; 257: 54-56Google Scholar, respectively. Actin was prepared as described by 31Spudich J.A. Watt S. J. Biol. Chem. 1971; 246: 4866-4873Google Scholar. Protein concentrations were determined by the absorption at 280 nm employing E1% of 5.5, 7.5, and 11.0 for myosin, S1, and actin, respectively, or by the 4Bradford M.M. Anal. Biochem. 1976; 72: 248-254Google Scholar using myosin or S1 as standards.Formation of Inhibitory Complexes in the Presence of NaF and ScCl3Trapping by the phosphate analogs was done by incubating S1 (10-20 μM) at 25°C in 0.03 M Tris, 10 mM NaF (or KF for trapping with beryllium or scandium) and 0.2 mM ADP at pH 7.9, adjusted at room temperature (Solvent A), with or without 1.0 mM MgCl2. After 5 min, the Pi analogs were added (up to 200 μM) and incubated for an additional 30 min. After incubation, an aliquot was diluted and stored on ice prior to determination of the NH4+/K+(EDTA) ATPase as described below. For experiments to determine the rates of inhibition by the Pi analogs, S1 was incubated as just described, and the reaction was started by addition of the Pi analogs to give a final concentration of 200 μM. Aliquots were removed and immediately assayed for ATPase activity. The rate of inhibition was also studied in the same buffer in the presence of 1.0 mM ATP by following the rates of Pi formation as described by 36White H.D. Methods Enzymol. 1982; 85B: 698-708Google Scholar.Determination of Scandium Trapped in the Inhibitory ComplexesThese were done using the inductively coupled plasma analysis by Galbraith Laboratories Inc.ATPase MeasurementsATPase assays were performed on the samples at 25°C in 1-ml aliquots containing 0.054 μM S1, 2 mM ATP, 0.56 M NH4Cl, 0.23 M KCl, 0.036 M EDTA, and 0.06 M Tris, pH 8.0. The amount of phosphate released was measured as reported by White(1982).Recovery of EDTA/ATPase Activity on Incubation with ActinThe stable S1 complexes were freed from excess reagents by sedimentation gel filtration (25Penefsky H.S. J. Biol. Chem. 1977; 252: 2891-2899Google Scholar) as described above in Sephadex G-50 equilibrated in 0.05 M Tris and 0.05 M KCl, pH 7.9 (adjusted at room temperature). F-actin was added to a final concentration of 10 μM, and the S1 complexes were 5 μM. The solutions were adjusted to 1.0 mM MgCl2. The samples were incubated at 25°C, and the release of ADP was estimated by the regain of the NH4+/K+(EDTA) ATPase at the time intervals indicated.Stability of the Inhibitory ComplexesThe stability of the complexes was followed by monitoring the release of trapped [14C]ADP upon addition of cold ADP as follows. [14C]ADP was trapped in S1 by incubation at 25°C in 0.05 M Tris, pH 7.9 (adjusted at 4°C), for 2 h with 10 mM NaF, 0.5 mM ScCl3, 5 mM MgCl2, 34.8 μM [14C]ADP, and S1 (17.4 μM). The untrapped reagents were removed by gel filtration sedimentation as described above and then for a second time in the gel equilibrated in 0.03 M Tris, 0.03 M NaCl, pH 7.9 (adjusted at 4°C). Both MgCl2 and ADP were added to 1.0 mM to the complex trapped with MgADP, and 1 mM ADP was added to the complex trapped in the absence of MgCl2. The release of [14C]ADP at 4°C with time was followed by determining the radioactivity and protein concentration after gel filtration sedimentation.Photocleavage of S1Photocleavage of the S1•ADP•ScFx and S1•MgADP•ScFx complexes was done in the presence of 0.5 or 1.0 mM Vi in 30 mM HEPES, pH 7.0, using the apparatus previously described (27Rajasekharan K.N. Sivaramakrishnan M. Burke M. J. Biol. Chem. 1987; 262: 11207-11214Google Scholar). Irradiations were done at 4°C for 10 min using a Pyrex Petri dish filter to eliminate far-UV.Gel ElectrophoresisSDS-polyacrylamide gel electrophoresis was done by the procedure of 18Laemmli U.K. Nature. 1970; 227: 680-685Google Scholar. The proteins were stained with Coomassie Brilliant Blue. A Xerox copy of the stained gel was made on Write-On Transparency Film (AF 4300, 3M Co.) yielding high quality images. Molecular weights were estimated with reference to the mobilities of reference proteins such as S1 and the light chains and the reported molecular weights for the Vi cleavage fragments of S1 (23Mocz G. Eur. J. Biochem. 1989; 179: 373-378Google Scholar; Werber et al., 1992).Fluorescence MeasurementsThese were done on a Perkin-Elmer LS-5B spectrofluorometer at 20°C. To S1 (0.9-1 μM) in 1 mM MgCl2, 30 mM Tris-HCl, 0.2 mM ADP, and 10 mM NaF, pH 7.9 (adjusted at room temperature), was added ScCl3 to a final concentration of 0.5 mM. The excitation wavelength was 295 nm, and the emission spectra were monitored between 310 and 390 nm.RESULTSIncubation of S1 at 25°C for 30 min in the presence of 0.2 mM ADP, 0.5 mM ScCl3, and 10 mM NaF in buffer A resulted in virtually complete inhibition of the EDTA ATPase of the protein. ADP and NaF were found to be essential for this inhibitory action since omission of either one had no effect on the ATPase activity of the S1. However, the presence of Mg2+ ions was not required to produce inhibition with ADP and fluoroscandium, except in this case lower extents of inhibition were observed at low concentrations of scandium (Fig. 1). The inactivation was studied as a function of the Sc3+ concentration, and the data are shown in Fig. 1in the presence of ADP (no MgCl2) and MgADP. In the latter case, 90% inhibition was observed with 0.1 mM scandium. The concentration dependence of the inhibition in the absence of Mg2+ ions is more complex and 90% inhibition required about 0.2 mM scandium. The loss of ATPase function is consistent with the formation of a stable inhibitory complex, and, since ADP, scandium, and NaF are all essential for this to occur, it is likely that this complex involves formation of a ternary complex in which MgADP or ADP and ScFx are simultaneously bound to the S1.The presence of MgATP in the incubation solution retards the onset of inhibition as shown in Fig. 2. This is consistent with the S1 being primarily in the MADP•Pi conformation and the occupancy of the site with MgADP and Pi with the latter retarding the binding of ScFx. The rate constant (kobs) for the inactivation in the presence of 500 μM scandium was obtained from the semilogarithmic plot of the residual ATPase activity versus time and has the value of 2.8 × 10-5 s−1, corresponding to an apparent second order rate constant (kapp) of 0.056 M−1 s−1. The latter second order rate constant is about an order of magnitude lower than that observed with Vi and beryllium suggesting weaker binding of the ScFx to S1•MgADP (Goodno, 1979; Phan and Reisler, 1992).Figure 2:Effect of the presence of MgATP (1.0 mM) on the rate of formation of Pi by S1 in the absence and presence of ScFx. To S1 (0.5 μM) at 25°C in the presence of 5.0 mM MgCl2, 1 mM ADP, 0.03 M HEPES, and 10 mM NaF at pH 7.0 (adjusted at room temperature) was added MgATP (1.0 mM), and the rate of Pi formation was followed in the absence (○) and presence (∆) of ScCl3 (500 μM).View Large Image Figure ViewerDownload (PPT)The inactivation of the ATPase activity induced by preincubation of S1 with MgADP in the presence of ScCl3 and KF was followed as a function of the preincubation time, and the data are shown in Fig. 3. The rate constant (kobs) for the inactivation in this case was obtained from the slope of the semilogarithmic plot of the residual activity versus time and was found to be 1.73 × 10-3 s−1 yielding an apparent second order rate constant of 8.7 M−1 s−1. While clearly faster than the apparent second order rate constant measured for the inactivation in the presence of MgATP, both rates are far too slow to correspond to a simple collisional process. It is more likely that the inactivation proceeds in a manner analogous to that suggested for Vi by Goodno(1979) and involves a rapid equilibrium binding step of ScFx to S1•MgADP followed by a slow isomerization step. Analyses of the amount of scandium trapped in the two complexes were done in triplicate on two different preparations of S1 and yielded 1.78 ± 0.3 and 1.61 ± 0.2 mol of scandium per mol of S1 for the inhibitory complexes containing MgADP and ADP, respectively. Since the amounts of nucleotide trapped in the S1 by the fluoroscandium anion were 0.72 ± 0.2 and 0.70 ± 0.1 for MgADP and ADP, respectively, per mol of S1, it appears that close to 2.0 mol of scandium per mol of S1 are present in these complexes. The amount of magnesium present in the complex formed with and without MgCl2 was 0.85 ± 0.02 and less than 0.14, respectively. S1 incubated with ScFx in the absence of MgCl2 and ADP showed less than 0.14 mol of scandium per mol of S1 after gel filtration centrifugation indicating that no high affinity binding site exists for ScFx when ADP is absent. Assuming that the total inactivation of the S1 ATPase is associated with the binding of 2.0 mol of Sc3+, a Scatchard analysis (data not shown) of the inactivation data versus free ScFx gave an association constant of approximately 6.0 × 104M−1. This is about an order of magnitude lower than that reported for the binding of BeFx to S1•MgADP (Phan and Reisler, 1992).Figure 3:Time dependence of inactivation of S1 by ScFx. S1 was incubated as described in Fig. 1(10 mM KF for beryllium and scandium), and the inactivation was followed as a function of time upon addition of ScCl3 (∇), BeCl2 (○), or Vi (∆) to a final concentration of 200 μM as described under “Materials and Methods.”◊, inactivation by ScCl3 in the absence of MgCl2.View Large Image Figure ViewerDownload (PPT)The stability of the inhibitory complexes formed in the presence or absence of MgCl2 was next examined by following the release of bound radiolabeled ADP by a chase experiment with 1.0 mM ADP. The results are shown in Fig. 4for the ternary complexes formed with either ADP or MgADP. The data for both processes indicate that the release is biphasic, with the fast phase having the same apparent rate constant (kobs) for both complexes of 8.0 × 10-5 s−1. However, the fractions of fast and slow components are quite different. In the case of the MgADP-containing complex, the fast component corresponds to about 0.46 of the total S1, while for the ADP-containing complex the fast fraction corresponds to only about 0.11 of the total S1. The slower phases yielded apparent rate constants of 8.0 × 10-6 s−1 and 3.9 × 10-6 s−1, corresponding to half-lives of 24 and 49 h for the MgADP- and ADP-containing inhibitory complexes, respectively, at 4°C.Figure 4:Rate of trapped [14C]ADP release. The inhibitory complexes formed with [14C]ADP and ScFx in the presence (○ and absence (∆) of MgCl2 were isolated at 4°C, and MgADP and ADP were added immediately to a final concentration of 1.0 mM respectively. The release of [14C]ADP was followed as described under “Materials and Methods.”View Large Image Figure ViewerDownload (PPT)Actin is known to be able to decompose the ternary inhibitory complexes formed between S1 and MgADP with Vi (Goodno and Taylor, 1982) or BeFx or AlFx (Phan and Reisler, 1992; Werber et al., 1992). The decomposition of the inhibitory complex is conveniently followed by the regain of the EDTA ATPase activity which is measured at high ionic strength where the activating effect by actin is minimal. The results of such an experiment with the ternary complexes formed with MgADP and ADP with ScFx and with MgADP and Vi (as a control) are presented in Fig. 5. It is clear that actin binding dissociates the ternary complex and results in the regain of the EDTA ATPase activity. The actin-induced rates of decomposition of these complexes can be roughly estimated from the half-life of regain of full ATPase activity. This gives a lower limit for the observed rate of complex decomposition (kobs), and the values were found to be 1.28 × 10-3 s−1 for both the MgADP-containing complexes trapped by Vi or ScFx. This is in reasonable agreement with the data of Werber et al.(1992) for the MMgADP•Vi complex under similar conditions. In contrast, the ADP-containing complex formed with ScFx is decomposed more slowly by actin at a rate of about 0.29 × 10-3 s−1.Figure 5:Effect of actin on the inhibitory complex. F-actin was added to a final concentration of 10 μM, and the S1 complexes were 5 μM. The solutions were adjusted to 1.0 mM MgCl2. The samples were incubated at 25°C, and the release of ADP was estimated by the regain of the NH4+/(EDTA) ATPase at the time intervals indicated. ∆, Vi; ○, ScFx(MgCl2); ◊, ScFx.View Large Image Figure ViewerDownload (PPT)It is well established that the tryptophyl fluorescence of S1 is sensitive to the state of the bound nucleotide, and, therefore, the intrinsic tryptophyl fluorescence of S1 in the presence of MgADP and NaF showed an enhancement of about 7.0% over that of S1 alone. Addition of Sc3+ caused an additional enhancement of 10% corresponding to a 17% enhancement of that for S1 in the absence of nucleotide. This is close to the value of 20% observed for the addition of MgATP to S1 and suggests that similar if not identical conformations exist for the MMgADP•ScFx and the MMgADP•Pi complexes. In the absence of MgCl2, addition of ADP to S1 in the presence of NaF resulted in a 2% enhancement of the tryptophyl fluorescence, which upon addition of the Sc3+ resulted in an additional enhancement of 14.6%, corresponding to a total enhancement of 16.6% over that of free S1. Thus, both the magnesium-containing and magnesium-free complexes stabilized by ScFx show very similar if not identical levels of intrinsic tryptophyl fluorescence.8Cremo C.R. Grammer J.C. Yount R.G. J. Biol. Chem. 1989; 264: 6608-6611Google Scholar have shown that irradiation of the MMgADP•Vi complex results in photo-oxidation of Ser-180 which is located in the Walker A consensus sequence of some ATP-requiring proteins. This result indicates that the bound Vi is located within chemical bond length of this seryl residue. In the absence of MgADP, irradiation of S1 in the presence of Vi results in cleavage at sites located at 23, 31, and 74 kDa from the N terminus with those at 23 and 31 kDa attributed to Vi binding to the same consensus sequence in S1 and a proximal serine, probably Ser-243 (28Ringel I. Peyser Y.M. Muhlrad A. Biochemistry. 1990; 29: 9091-9096Google Scholar). These cleavages give rise to bands of 74, 51, 46, and 23 kDa in SDS gel electrophoretograms. Addition of Vi to the ternary complexes formed with ScFx, and subsequent irradiation by near-UV light showed cleavage only at the 74-kDa site (Fig. 6). The failure to observe bands of 31 and 23 kDa in the two forms of ScFx complexes upon irradiation in the presence of Vi suggests that Vi is unable to bind to the consensus sequence site due to the occupancy by the ScFx anion.Figure 6:SDS-polyacrylamide gel electrophoresis analyses of the effect of Vi-induced photolysis of S1 and of the inhibitory complexes formed with ScFx and MgADP or ADP. Lanes: a, S1 with Vi (1.0 mM); b, S1•MgADP•ScFx with Vi (0.5 mM); c, same as b except Vi was 1.0 mM; d, S1•DP•ScFxwith Vi (0.5 mM); e, same as d except Vi was 1.0 mM; f, control S1.View Large Image Figure ViewerDownload (PPT)DISCUSSIONThe present study was undertaken to examine whether fluoride complexes of a transition metal cation such as scandium could also mimic BeFx and AlFx in trapping MgADP in S1 to form a stable complex at low temperature. The reasons for choosing scandium were as follows. (i) The stability of the fluoroscandium complexes is similar to fluoroaluminates (11Goldstein G. Anal. Chem. 1964; 36: 243-244Google Scholar) which have already been used to trap MgADP in S1. (ii) The exchange rates of ligands for scandium lie between those of beryllium and aluminum ions (32van Eldik R. Inorganic High Pressure Chemistry: Kinetics & Mechanisms. Elsevier Science Publishing Co. Inc., New York1986Google Scholar; 37Wilkins R.G. Kinetics and Mechanism of Reactions of Transition Metal Complexes. VCH Publishers, New York, 2nd Ed1991Google Scholar). (iii) We were interested to determine whether the hexacoordinated scandium anions, which may be in a deformed octahedral configuration (24Nowacki W. Z. Kristallogr. 1939; 101: 273-283Google Scholar), would be effective in the protein to trap MgADP or free ADP. The data presented in Fig. 1and Fig. 2indicate that fluoroscandium anion (presumably ScF4¯) can indeed produce an inhibitory complex of S1 in the presence of MgADP. Surprisingly, inhibition was also observed when Mg2+ ions were omitted from the incubation mixture, indicating that ADP can also be trapped in S1 by fluoroscandium. Direct chemical analyses of the total amount of magnesium in solutions of S1 (13 μM) in the absence of MgCl2 yielded a maximum of 3.2 μM magnesium content, while in solutions of the purified complex (at 13 μM) formed in the absence of magnesium, the concentration of total magnesium was less than 2.0 μM which would yield a total concentration of less than 0.5 μM in MgADP in these solutions which is far too low to account for stoichiometric trapping of MgADP in the protein.Analysis for the amount of scandium present in these complexes indicate that 2 mol of scandium are present per mol of S1. The reason for this is not as yet apparent. It is, however, possible to rule out a high affinity site for the binding of ScFx in the absence of ADP, since little binding was observed when ADP and MgCl2 were absent (see “Materials and Methods”). It is unlikely that any of the bound scandium is in the form of Sc3+ based on the association constants for the formation of the fluoroscandium species in the presence of the high concentrations of NaF employed (Goldstein, 1964). Therefore, it is highly improbable that the complex formed in the absence of MgCl2 involves any substitution of Sc3+ for Mg2+ in coordinating to the bound ADP. We have also found that inhibition occurs with BeFx and AlFx with ADP in the absence of free Mg2+, but much higher concentrations of the fluorometal anions are needed. 2D. Gopal and M. Burke, unpublished data. The apparent second order rate constants (kapp) for the onset of inhibition, 0.056 M−1 s−1 in the presence of MgATP and 8.7 M−1 s−1 for preincubation in its absence, are far too slow for a simple collisional process of ScFx with S1•MgADP and suggest that a very slow rate-limiting isomerization step occurs after formation of the collisional complex as originally suggested by Goodno (1979) for the Vi-induced inhibition of S1 in the presence of MgADP.The stability of the ternary complexes formed with fluoroscandium anion and MgADP or ADP at 4°C was followed by the release of the trapped 14C-nucleotide by chasing the isolated complexes with cold ADP as described in Fig. 4. The time dependence of the release indicated that it was biphasic for both forms of the complex. A similar biphasic release was also observed with the complex stabilized with BeFx (Phan and Reisler, 1992). For the two ScFxcomplexes, the fast phase appeared to be identical (kobs = 8.0 × 10-5 s−1) but the fractions of S1 associated with the fast phase were different being 0.46 and 0.11 for the MgADP-containing and ADP-containing complexes, respectively. The rates of the slow phases were also found to be different for these complexes. The biphasic nature of the release of nucleotide from these ternary complexes suggests the possibility that the ternary complex may be in equilibrium between two distinct conformations, one in which the nucleotide can be released more rapidly and a form where the nucleotide is more strongly trapped. A similar mechanism has been proposed by Mihashi et al.,(1990) for the MMgADP•Vi complex. In that study it was shown, using an exchange procedure, that nucleotide could undergo exchange with trapping, suggesting that release of nucleotide from the inhibitory equilibrium mixture preceded release of the Vi anion. Furthermore, the rate constant of the rapid phase for the exchange in the case of the Vi-stabilized complexes was 4.5 × 10-4 s−1 which is about a factor of 6 times greater than that observed for the release of ADP from the ScFx-trapped MgADP in S1 measured at 4°C. The rates of the slow phase of releases of ADP from the MgADP-containing ScFx and ADP-containing complexes, 8.0 × 10-6 s−1 and 3.9 × 10-6 s−1, respectively, are in reasonable accord with the slow rate of exchange measured for the MMgADP•Vi complex (6.7 × 10-6 s−1) (Mihashi et al., 1990).The enhancement of tryptophyl fluorescence associated with the formation of the two complexes (presence and absence of MgCl2) is quite similar, allowing for experimental error, and shows levels of enhancement over that of unliganded S1 of about 17 ± 2%. This is somewhat lower than that observed for the MMgADP•Pi complex (20%± 2%) but suggests that the conformations of the complexes may be closely related to this steady-state intermediate. The failure to detect any photocleavage at the two sites (21 and 31 kDa from the N terminus) in the presence of added Vi for the two complexes indicates that Vi probably does not bind to the region of S1 containing the ATP consensus sequence (33Walker J.E. Saraste M. Runswick M.J. Gay N.J. EMBO J. 1982; 1: 945-951Google Scholar), since it is thought that occupancy of this region by Vi is essential for the photocleavage to occur (Werber et al., 1992). This observation is essentially identical with that noted for the complexes stabilized with BeFx and AlFx, respectively, and would be consistent with the ScFx occupying this site in the inhibitory complexes and the inability of the added Vi to displace it (Werber et al., 1992).Perhaps the main question concerning these inhibitory complexes stabilized by the various Pi analogs is to what extent do they mimic the MMgADP•Pi state. The large enhancement of the tryptophyl fluorescence accompanying their formation and the weak actin binding are in accord with their attaining a structure similar to this steady-state intermediate. However, these correlations may be coincidental. Another possibility is that, while the initial collisional complex may be related to the MMgADP•Pi state, the much slower isomerization step leading to inactivation may involve a rearrangement of the ligands in the protein. For example, there is some evidence that ADP may interact weakly with BeFxeven in the presence of Mg2+ (17Issartel J.P. Dupuis A. Lunardi J. Vignais P.V. Biochemistry. 1991; 30: 4726-4733Google Scholar). While formation of such a complex may be poorly favored in free solution, it is conceivable that in the protein such a complex may be stabilized. While it is usually assumed that the ligands bound in these inhibitory complexes are present in a fashion analogous to those in the MMgADP•Pi complex, that is they are in the M•MgADP•ScFx state, there is no direct evidence to our knowledge that can rule out the possibility that the slow isomerization leading to inhibition does not form a complex of the form M•ScFx•MgADP in which the nucleotide is locked not at the active site but at another site due to a rearrangement. In such a situation the complex would still be inhibitory with the Pi analog blocking access to the ATPase site. Some support for this view is the recent finding that photoaffinity-labeled S1, formed by irradiation of complexes trapped in S1 by Vi, containing covalently linked nucleotide analog show normal levels of ATPase, suggesting that the catalytic site was not labeled in these complexes (20Luo Y. Wang D. Cremo C. Pate E. Cooke R. Yount R. Biophys. J. 1994; 66: A79Google Scholar). 3R. Yount, personal communication. Secondly the ability of the Vi-stabilized complexes to undergo nucleotide exchange with trapping indicates that nucleotide release is faster than that of Vi (Mihashi et al., 1990), and this is exactly the opposite order of release occurring in the normal hydrolytic cycle with MgATP (1Bagshaw C.R. Trentham D.R. Biochem. J. 1973; 133: 323-328Google Scholar, 2Bagshaw C.R. Trentham D.R. Biochem. J. 1974; 141: 331-349Google Scholar).In summary, the evidence presented in this paper suggests that ScFx is capable of trapping either MgADP or ADP in S1 resulting in the formation of rather stable inhibitory complexes (at 4°C) and extends the range of Pi-like anionic analogs that are able to function in this manner. Perhaps the most surprising outcome of the present work is that free ADP is also apparently effectively trapped by this anion in S1. The two complexes, with MgADP or ADP, show clear differences in their stabilities at 4°C and at 25°C in the presence of actin, and, in fact, the complex formed with ADP appears to have the higher stability. INTRODUCTIONThe energy for muscle contraction is derived from the hydrolysis of MgATP by myosin and the activation of this process by actin. During this process, myosin forms a series of intermediates resulting in the formation of a steady-state intermediate MADP•Pi which upon reassociation with actin forms AMADP•Pi, a non-force-producing state, which subsequently isomerizes to force-bearing states with and without Pi (AMADP•Pi or AMADP), respectively (16Hibberd M.G. Trentham D.R. Annu. Rev. Biophys. Biophys. Chem. 1986; 15: 119-161Google Scholar; 9Dantzig J.A. Goldman Y.E. Millar N.C. Lacktis J. Homsher E. J. Physiol. 1992; 451: 247-248Google Scholar). Thus, a key question related to the structural basis of force generation is how different is the S1 1The abbreviations used are: S1myosin subfragment 1VivanadateBeFAlF, and ScF are generalized representations of fluorometal complexes formed with beryllium, aluminum, and scandium. structure in the MADP•Pi and MADP•Pi or MADP states. Unfortunately, the former two intermediates have only limited lifetimes, and it appears that the latter state when bound to actin is different in structure from that formed by the binding of MgADP to actoS1 (29Sleep J.A. Hutton R.L. Biochemistry. 1980; 19: 1276-1283Google Scholar). Therefore, stable analogs corresponding to these species would be useful in obtaining information of their structures.Since the demonstration by Goodno(1979) that orthovanadate ion in the presence of MgADP was able to form a stable ATPase-inhibitory complex with myosin S1, corresponding to a 1:1:1 ternary complex designated MADP•Vi, much interest has been devoted in its characterization as it appears to have some properties in common with the steady-state intermediate, MADP•Pi, of the MgATPase hydrolysis cycle by S1 (13Goodno C.C. Methods Enzymol. 1982; 85B: 116-123Google Scholar; 30Smith S.J. Eisenberg E. Eur. J. Biochem. 1990; 193: 69-73Google Scholar; see also 14Goodno C.C. Taylor E.R.W. Proc. Natl. Acad. Sci. U. S. A. 1982; 79: 21-25Google Scholar for a review)). Because of the simple stoichiometry of the complex and because of its inhibitory nature, it is generally assumed that the nucleoside diphosphate resides at the nucleotide binding site with the Vi taking the place of the terminal γ Pi group and, in this way, forms a stable analog of the transition state. Such an interpretation is also supported by the fact that structurally the vanadate ion is similar to orthophosphate (19Lindquist R.N. Lymn J.L. Lienhard G.E. J. Am. Chem. Soc. 1973; 95: 8762-8768Google Scholar).More recently, the use of other complex anions with similarity to Pi has been investigated for their ability to stabilize the transition states of phosphotransferases and nucleotidases in the presence of nucleotide diphosphate (5Chabre M. Trends Biochem. Sci. 1990; 15: 6-10Google Scholar). These include fluoride complexes of beryllium and aluminum, BeFx and AlFx, respectively, and their efficacy in forming stable transition state analog states has been studied with actin (6Combeau C. Carlier M.-F. J. Biol. Chem. 1989; 264: 19017-19021Google Scholar, 1990), with smooth muscle heavy meromyosin (21Maruta S. Henry G.D. Sykes B.D. Ikebe M. J. Biol. Chem. 1993; 268: 7093-7100Google Scholar) and skeletal myosin S1 (26Phan B. Reisler E. Biochemistry. 1992; 31: 4787-4793Google Scholar; 3Beck T.W. Chase P.B. Kushmerick M.J. Biophys. J. 1992; 61: 2506aGoogle Scholar: 35Werber M.M. Peyser Y.M. Muhlrad A. Biochemistry. 1992; 31: 7190-7197Google Scholar; 15Henry G.D. Maruta S. Ikebe M. Sykes B.D. Biochemistry. 1993; 32: 10451-10456Google Scholar).In the present paper we present evidence that fluoroscandium anion (presumably ScFx) can also lead to effective trapping of ADP in S1 in both the presence and absence of Mg2+ with concomitant inactivation of the ATPase properties of the protein. Surprisingly, free nucleoside diphosphate is more stably trapped than the MgADP form with fluoroscandium. Actin binding to these complexes results in a faster release of trapped ADP in accord with the usual effect of actin in accelerating product release from S1.