Báo cáo Y học: Effect of adenosine 5¢-[b,c-imido]triphosphate on myosin head domain movements Saturation transfer EPR measurements without low-power phase setting ppt

Effect of adenosine 5¢-[b,c-imido]triphosphate on myosin head domain movements Saturation transfer EPR measurements without low-power phase setting No ´ ra Hartvig 1 ,De ´ nes Lo ˜ rinczy 2 , Nelli Farkas 1 and Joseph Belagyi 1 1 Central Research Laboratory and 2 Institute of Biophysics, School of Medicine, University of Pe ´ cs, Hungary Conventional and saturation transfer electron paramagnetic resonance spectroscopy (EPR and ST EPR) was used to study the orientation of probe molecules in muscle fibers in different intermediate states of the ATP hydrolysis cycle. A separate procedure was used to obtain ST EPR spectra with precise phase settings even in the case of samples with low spectral intensity. Fibers prepared from rabbit psoas muscle were labeled with isothiocyanate spin labels at the reactive thiol sites of the catalytic domain of myosin. In comparison with rigor, a significant difference was detected in the orientation- dependence of spin labels in the ADP and adenosine 5¢-[b,c-imido]triphosphate (AdoPP[CH 2 ]P) states, indica- ting changes in the internal dynamics and domain orienta- tion of myosin. In the AdoPP[CH 2 ]P state, approximately half of the myosin heads reflected the motional state of ADP–myosin, and the other half showed a different dynamic state with greater mobility. Keywords: adenosine 5¢-[b, c-imido]triphosphate (AdoPP- [CH 2 ]P); myosin; saturation transfer EPR; spin-labeling. It is generally accepted that domain movements in the myosin head play a decisive role in the energy-transduction process of muscle contraction [1–5]. It is a multistep process which can produce several conformational states of myosin [6–9]. Extensive studies using different techniques have indicated that the nucleotide-binding pocket does not experience large conformational changes during the hydro- lytic cycle [10–12]. However, small conformational changes induced by nucleotides in the motor domain should be converted into larger movements. The data show that, whereas the structure of the motor domain remains similar to rigor, the regulatory domain swings around a point at the distal end of the motor domain [8,13–17]. The changes in the 50-kDa domain affect the segment of the 20-kDa domain that contains the essential thiol groups, SH1 (Cys707) and SH2 (Cys697). This part may be involved in the transducing of small conformational changes [18,21,22]. Spectroscopic probes are widely used in muscle research. Paramagnetic probes provide a direct method by which dynamic changes and the rotation and orien- tation of specifically labeled proteins can be followed. In muscle fiber studies, the probe molecules, particularly the maleimide-based nitroxides and iodoacetamide spin labels, are usually attached to the reactive thiol site Cys707 of the motor domain [21–23], or to the regulatory light chain [9,24,25]. The main problems that limit interpretation of spectroscopic measurements are the location of the probe molecules on the proteins, the relative orientation of the spin labels with respect to the magnetic field, and the perturbing effect of probes on structure and function. We have previously observed that an isothiocyanate-based spin label is more sensitive to the domain orientation in the myosin head than the widely used maleimide spin label [26]. Selective modification of Cys707 strongly affects MgATP hydrolysis, and the motor function of myosin heads in the in vitro motility assay is blocked [27,28]. The most pronounced effect was observed in the intermediate complex ADP–P i [19,20]. Spin label molecules bound to myosin are able to detect nucleotide binding and conformational changes in the myosin head related to the hydrolytic cycle of ATP [29]: AM þ ATP $ A þ MÁATP $ A þ M à ÁATP $ AM Ãà ÁADPÁP i $ AM à ÁADP þ P i $ AM þ ADP þ P i where M denotes myosin, A stands for actin, and the asterisk (*) distinguishes intermediate conformational chan- ges. From consideration of the crystal structure of the myosin head [3,4], and from fluorescence measurements, it was concluded that the so-called open-closed transition does not require hydrolysis of ATP [30]. It was also suggested that the nonhydrolyzable analogue adenosine 5¢-[b,c-imi- do]triphosphate (AdoPP[CH 2 ]P) was able to induce the transition. The intermediate states have different spectral properties, and distinct conformations can be assigned to these states. Our primary aim was to use the spin label EPR technique to measure changes in label orientation and/or rotational rate that results from binding of AdoPP[CH 2 ]P to myosin in muscle fibers, and to find some correlation with the myosin head structure. The AdoPP[CH 2 ]P state can be stabilized for long enough to conduct measurements by Correspondence to J. Belagyi, Central Research Laboratory, Univer- sity of Pe ´ cs School of Medicine, 12 Szigeti Str., H-7643 Pe ´ cs, Hungary. Fax: + 36 72 536254, Tel.: + 36 72 536255, E-mail: belagyi@apacs.pote.hu Abbreviations:AdoPP[CH 2 ]P, adenosine 5¢-[b,c-imido]triphosphate; ST EPR, saturation transfer electron paramagnetic resonance spectroscopy; MSL, 4-maleimido-2,2,6,6-tetramethylpiperidino-oxyl; TCSL, 4-isothiocyanato-2,2,6,6-tetramethylpiperidino-oxyl; TNBS, 2,4,6-trinitrobenzenesulfonate. (Received 7 August 2001, revised 4 March 2002, accepted 7 March 2002) Eur. J. Biochem. 269, 2168–2177 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02872.x saturation transfer (ST) EPR. ST EPR is widely applied to biological problems [9,10,12,21]. Motion of the label is slow in biological samples, and the spectral intensity is not too high, therefore this problem motivated us to find a way to eliminate the critical setting of the out-of-phase null at unfavorable signal-to-noise ratio. MATERIALS AND METHODS Materials KCl, MgCl 2 , EGTA, histidine/HCl, glycerol, ADP, ATP, AdoPP[CH 2 ]P, N-ethylmaleimide, 4-maleimido-2,2,6, 6-tetramethylpiperidino-oxyl (MSL) and 4-isothiocyanato- 2,2,6,6-tetramethanepiperidino-oxyl (TCSL), phosphoenol- pyruvic acid, pyruvate kinase, and lactate dehydrogenase were obtained from Sigma (Germany). 2,4,6-Trinitroben- zenesulfonate (TNBS) was obtained from Fluka (Germany). Fiber preparation Glycerol-extracted muscle fiber bundles were prepared from rabbit psoas muscle. Small strips of muscle fibers were stored after osmotic shock in buffer containing 50% (v/v) glycerol, 80 m M KCl, 5 m M MgCl 2 ,1m M EGTA and 25 m M Tris/HCl, pH 7.0, at )18 °C for up to 1 month. Fiber bundles from glycerinated muscle were washed for 60 min in rigor buffer (80 m M potassium propionate, 5 m M MgCl 2 ,1m M EGTA, 25 m M Tris/HCl buffer, pH 7.0), which removed glycerol, and then transferred to fresh buffer. This state models the rigor state of the muscle. We added MgADP to the rigor solution at a final concentration of 5 m M to simulate the strongly binding state of myosin to actin, which may correspond to the AM–ADP state. In experiments involving MgADP, the activity of adenylate kinase was inhibited by the addition of 50 l M diadenosine pentaphosphate. The other analogue of intermediates in the ATPase pathway was formed by AdoPP[CH 2 ]P,which stoichiometrically binds at the active site of myosin to form a stable complex. The muscle fibers were stored in solution containing 80 m M potassium propionate, 5 m M MgCl 2 , 1m M EGTA, 5 or 16 m M AdoPP[CH 2 ]P and 25 m M Tris/ HCl, pH 7.0, for 15 min at 0 °C,andthenspectrawere recorded at ambient temperature (20–22 °C). Only freshly opened bottles of AdoPP[CH 2 ]P were used to avoid degradation of the compound [31]. Preparation of MSL–hemoglobin Freshly prepared human hemoglobin was spin-labeled with MSL for 24 h; the molar ratio of label and protein was one to one. The labeled site was the b-93 cysteine residue. The spin-labeled protein was dialyzed against 0.1 M phosphate buffer at pH 7.0 for 48 h at 4 °C and lyophilized. Then 5 mg protein was dissolved in 1 mL buffer (final concen- tration) to obtain samples. Different amounts of glycerol up to 70% (v/v) and sucrose up to 50% (w/v) were added to increase the viscosity. Spin labeling of muscle fibers Spin labeling of fibers was performed in labeling medium (rigor solution plus 2 m M pyrophosphate at pH 6.5) with % 2 mol TCSL to 1 mol myosin for 20 min at 0 °C. Before spin labeling, the fibers were incubated in low-ionic strength buffer [1 m M EGTA, 5 m M MgCl 2 ,1m M 5,5¢-dithiobis(2- nitrobenzoic acid) and 20 m M Mops, pH 7.5] for 1 h to achieve selective labeling of the reactive thiols [32]. After spin labeling, the fiber bundles were washed in rigor buffer plus 5 m M dithiothreitol for 30 min at 0 °C, pH 7.0, to remove the unreacted labels and restore the preblocked thiol groups. The ratio of the attached spin labels to moles of myosin varied between 0.22 and 0.41 (mean value 0.33) as calculated from the double integral of the EPR spectra after comparison with the double integral of an MSL solution of 10 l M using the same sample cell and spectrometer parameters. EPR measurements Conventional and ST EPR spectra were recorded with an ESP 300E (Bruker) spectrometer. First harmonic, in-phase, absorption spectra were obtained by using 20 mW micro- wave power and 100 kHz field modulation with amplitude of 0.1 or 0.2 mT. Second harmonic, 90° out-of-phase, absorption spectra were recorded with 63 mW and 50 kHz field modulation of 0.5 mT amplitude detecting the signals at 100 kHz out-of-phase. The 63 mW microwave power corresponds to an average microwave field amplitude of 0.025 mT in the central region of the standard tissue cell of Zeiss (Carl Zeiss), and the values were obtained by using the standard protocol of Fajer & Marsh [33]. In this region of the tissue cell, small segments of the muscle fibers (6–7 mm long) were mounted parallel to each other. The spectra were recorded in two positions at a temperature of 22 ± 1 °C, where the longer axis of the fibers was oriented parallel and perpendicular to the laboratory field. The spectra were normalized to the same number of unpaired electrons by calculating the double integral of the derived spectra. We assumed that the spectra from TCSL-fibers in different states would be composed of a linear combination of spectra; normalized EPR spectra were manipulated by digital subtraction. To obtain the precise phase setting for the ST EPR technique, a different procedure was applied [34,35]. The idea originates from B. H. Robinson (Department of Chemistry, Nashville University, Nashville, KS, USA). Assuming that at low microwave power the variance of the EPR signal would be minimum over the whole field scan at the out-of-phase setting, the correct phase angle can be calculated from two high-power spectra differing in phase angle by exactly 90°. Using the least-squares assumption at the out-of-phase setting for an EPR spectrum, we obtain S ¼ X n i ½spðiÞÀm 2 ¼ min where sp(i) is the signal intensity of the digitized spectrum, and m is the mean value of the EPR signal. The summation variable i ranges from 1 to n,wheren denotes the number of digitized data. In general the spectrum at an arbitrary phase setting is the linear combination of the in-phase and out-of- phase components spðiÞ¼AðiÞ cos 0 þ BðiÞ sin 0 Ó FEBS 2002 Effect of AdoPP[CH 2 ]P on myosin head domain movements (Eur. J. Biochem. 269) 2169 where J is the phase angle, and A(i)andB(i)arethe amplitudes of the in-phase and out-of-phase components. Differentiating S with respect to variables m and J,and setting the derivatives equal to zero results in m ¼ spðiÞ and tan 20 ¼ 2V AB V B À V A where V A , V B are the variances of the components A(i)and B(i), and V AB is the common variance. To apply this procedure to a given sample and to obtain the in-phase and out-of-phase spectra, two independent data sets (two spectra) are required for the same sample, for which the phase angles differ by exactly 90°: sp1ðiÞ¼AðiÞ cos 0 þ BðiÞ sin 0 sp2ðiÞ¼ÀAðiÞ sin 0 þ BðiÞ cos 0 where sp1(i) is taken at an arbitrary phase angle J,and sp2(i) is recorded at a phase angle J +90°.Linear transformation of the two spectra through the calculated phase angle J allows the second harmonic in-phase and out- of-phase EPR spectra to be estimated. In practice, the digitized data of sp1(i)andsp2(i) are used to obtain the variances and the phase angle. The linear transformation through the calculated phase angle gives the required out- of-phase spectrum. The upper spectra in Fig. 1 are two second-harmonic high-power EPR spectra for MSL-hae- moglobin at phase angles a and a +90°, and the lower spectrum is the out-of-phase spectrum calculated from the new phase angle. In control experiments, the 0-degree was obtained by adjusting the phase angle by the method of Squier & Thomas [36]. In this paper we use the expression method of variance to distinguish this procedure from the conventionally used low-power phase setting (null method). The procedure was tested on MSL-haemoglobin in different regions of the rotational correlation times. The viscosity of the samples was increased by the addition of sucrose or glycerol. The temperature of the samples was varied in the range 20 °Cto)30 °C by using an ER412 VT temperature regulator from Bruker. In other cases, skeletal muscle F-actin and glycerinated muscle fibers labeled with N-maleimide or isothiocyanate spin labels were measured. ATPase activity ATPase activity was determined using a pyruvate kinase/ lactate dehydrogenase-coupled optical test [29]. The assay medium for Mg 2+ -ATPase consisted of 100 m M KCl, 20 m M Mops, 1 m M MgCl 2 ,0.5m M EGTA, 0.15 m M NADH, 1 m M phosphoenolpyruvate, 20 UÆmL )1 pyru- vate kinase, 40 UÆmL )1 lactate dehydrogenase, and 0.5 m M ATP, pH 7.0. For Ca 2+ ,Mg 2+ -ATPase, the assay medium also contained 1 m M CaCl 2 . At 340 nm, the absorption change was measured with a Perkin–Elmer spectrophotometer interfaced to a computer. The molar absorption coefficient (e 340 )ofNADHwas 6.22 · 10 3 mol )1 Æcm )1 . In the experiments, the Mg 2+ - ATPase and Ca 2+ ,Mg 2+ -ATPase activities of thin fiber bundles over 10-min intervals were determined. Fiber bundles of 8–10 mg wet weight were slightly stretched on a rectangular support made of platinum. The support was diagonally fitted into a standard quartz cuvette, which was filled in with the solutions. The solution was continuously mixed with a small magnetic bar. The decrease in absorbance resulted in a straight line, and the slope of the straight line was used to estimate ATPase activity. We assumed that 50% of the dried muscle weight was myosin. The Mg 2+ -ATPase activity was 4.131 ± 0.718 lmol P i Æ(mg myosin) )1 Æmin )1 (n ¼ 4) for control fibers and 4.024 ± 0.742 lmol P i Æ(mg myo- sin) )1 Æmin )1 (n ¼ 4) for TCSL-fibers. The ATPase activity of active fiber bundles was 5.565 ± 0.816 lmol P i Æ(mg myosin) )1 Æmin )1 (n ¼ 4) for control fibers, and 3.199 ± 0.457 lmol P i Æ(mg myosin) )1 Æmin )1 (n ¼ 4) was calculated for TCSL-fibers. The K + /EDTA ATPase activity of myosin extracted from muscle fibers was measured by determining the rate of release of P i as described previously [26]. The mean ATPase activity of myosin was reduced to % 60% of the control after spin labeling with TCSL. Fig. 1. Component spectra of MSL-hemoglobin (A) and ST EPR spectrum of MSL-hemoglobin calculated from the two high-power component spectra (B). The phase angles of the spectra in (A) differ by 90°. 2170 N. Hartvig et al.(Eur. J. Biochem. 269) Ó FEBS 2002 RESULTS Rotational dynamics of spin-labeled hemoglobin Figures 2 and 3 show the comparison of the ratio of the diagnostic peaks L¢¢/L and the EPR spectra. The MSL- haemoglobin samples were subsequently measured with the variance method and the conventional null technique [36]. Appropriate fitting provides evidence that the variance method is useful, especially when the rotational motion of large proteins with low concentration of spin labels should be detected. At low concentration of spin labels where high receiver gain is required to obtain a spectrum of good quality, it is difficult to follow the widely used low-power (£ 1 mW) phase setting method because of the unfavorable signal-to-noise ratio. In a series (n ¼ 9) of spin-labeled muscle fibers, the phase setting adjusted by the null method (u ¼ 0°)was changed by exactly 40°. The spectra were then recorded, and the phase angle was calculated by the variance technique on each sample. The result of the measurements was u ¼ 42.38 ± 2.09 (mean ± SD). Good agreement was obtained with the integral method [37,38] (Fig. 4). On the basis of the comparison, we suggest that the variance method is almost equivalent to the other methods. Orientation of probe molecules Myosin in fibers was spin-labeled with an isothiocyanate- basedspinlabel,whichisbelievedtobindtothefast reacting thiol sites in the catalytic domain of myosin. The labels in the fibers were immobilized on the microsecond time scale: the effective rotational correlation time was 60 ls in the absence of nucleotides calculated from ST EPR spectra (Fig. 5C). Spectroscopic probes provided direct information about the orientation of the myosin heads; in rigor, they had only one mode of binding [22,23]. The MSL probes showed a narrow distribution with respect to the longer axis of the fibers, with a mean angle of 82° and an angular spread of 6°. With TCSL probes, the EPR spectra also reported a high dependence of orientation [26], but with different mean angle and angular spread (J ¼ 75°, r ¼ 16°, Fig. 4) compared with MSL-fibers. Blocking the reactive thiol site, Cys707, with 0.1 m M N-ethylmale- imide before spin labeling, but after incubation with 5,5¢-dithiobis(2-nitrobenzoic acid), greatly reduced the spec- tral intensity of the TCSL-fibers. Using the same procedure of TCSL labeling without and with N-ethylmaleimide, the molar ratio of the bound spin label to mol of protein was about 25 : 1. After N-ethylmaleimide pretreatment, the spectrum showed a fraction of weakly immobilized labels, and the orientation-dependence of TCSL-fibers was greatly reduced. Pretreating the fibers with TNBS before Fig. 2. Representative ST EPR spectra recorded for MSL-hemoglobin. Lyophilized MSL-hemoglobin was dissolved in 0.1 M phosphate buf- fer, pH 7.0, and either glycerol or sucrose plus glycerol was added to the samples to increase viscosity. The final concentration of hemo- globin was 5 mgÆmL )1 . Upper curve: null method; lower curve: vari- ance method. (A) MSL-hemoglobin in 70% (v/v) glycerol at 0 °C; (B) MSL-hemoglobin in 70% (v/v) glycerol plus 30% (w/v) sucrose at 0 °C; (C) MSL-hemoglobin in 70% (v/v) glycerol plus 50% (w/v) sucrose at )30 °C; (D) lyophilized MSL-hemoglobin. Fig. 3. Comparison of the spectroscopic methods using the low-field diagnostic peaks (L¢¢/L) on different spin-labeled samples. For the variance method the phase angle was calculated from two high-power out-of-phase EPR spectra. The regression coefficient of the fitted straight line is shown. Ó FEBS 2002 Effect of AdoPP[CH 2 ]P on myosin head domain movements (Eur. J. Biochem. 269) 2171 spin-labeling did not significantly affect the orientation- dependence of the EPR spectra compared with that of fibers pretreated with 5,5¢-dithiobis(2-nitrobenzoic acid). On the other hand, trinitrophenylation of the reactive lysine residue before TCSL treatment did not significantly modify the orientation-dependence of the labels derived from the spectra. Spectrum simulation based on our earlier work [26] showed that, although the TCSL probe molecules were probably attached to the same sites as the MSL probe molecules, they exhibited different orientational distribu- tion. To determine the orientation of the probe molecule within the myosin head, the electron microscopic data of the actomyosin complex [39,40], the crystal structure of sub- fragment 1 [4], and the procedure developed by Fajer [23] were taken into account. The average angle of attachment of the myosin head to the actin filament was % 40° at practically all stages of the hydrolytic cycle of ATP [39]; only a few degrees of difference were detected in the tilt angle. It suggests that, in the presence of ADP, the TCSL label reflects an internal rearrangement of the structure in the environment of the labeled site. Accepting the reference system for the head of myosin defined by Fajer [23], the x axis of the reference frame lies in the plane of the head projection and inclined at 40° with respect to the long axis of the fibres, while the y-axis is perpendicular to this plane. Then to a rough approximation, the tilt angle of the principal z-axis of the TCSL label is about 30°,using the data derived from our model-independent approach for the calculation of orientational distribution. A significant change in the tilt angle of the principal z axis after addition of ADP or AdoPP[CH 2 ]P would mean reorientation of the segment, i.e. the broken helix containing the two reactive thiols, that holds the label. The uncertainty arising from the incorrect orientation of the fibre segments with respect to Fig. 5. EPR spectra of TCSL-fibers in rigor and the ADP state. Spectra A and B were recorded in parallel orientation, and spectrum C was obtained in perpendicular orientation to the fibers. The fibers were kept in buffer solution (see Materials and methods) during spectrum accumulation. MgADP was added at a concentration of 5 m M ,and the fibers were kept for 10 min in buffer before EPR measurement. The change in the hyperfine splitting (2A¢ zz ) shows the different static orders of spin labels in rigor and the ADP state. (A) Rigor state; (B) ADP state; (C) rigor state; (D) ST EPR spectrum of rigor fibers. ST EPR spectra were recorded for fibers oriented perpendicular to the long axis of the fibers. The field scan was 10 mT. Fig. 4. Comparison of the two methods by calculating the first integrals of the corresponding two spectra. The regression coefficient of the fitted straight line is shown. 2172 N. Hartvig et al.(Eur. J. Biochem. 269) Ó FEBS 2002 the magnetic field and from the assumption of an axially symmetrical EPR parameter A in the model-independent approach of the calculation has an impact on the values for probe orientation. Effect of nucleotides The addition of MgADP resulted in a change in the mean angle of the distribution of the spin labels. It decreased from 75° to 56°, and the angular spread increased by 4°,butthe orientation order remained (Fig. 5). In contrast, AdoPP[CH 2 ]P produced orientation disorder of the myosin heads, and a random population of spin labels was superimposed on the ADP-like spectrum, which demon- strates conformational/motional changes and dissociation of the myosin heads (Fig. 6). In agreement with previous data, the fraction of the ordered population was estimated to be % 50% of the total concentration [41]. However, there was a large difference between the spectra of MSL-fibers and TCSL-fibers. One component of the spectra of MSL- fibers reflected exactly the same orientation as in rigor; the myosin heads apparently remained bound to actin. The second component was characteristic of randomly oriented myosin heads (Fig. 7). In the case of the TCSL-fibers, the component of the spectrum with a high degree of order was the same as in the ADP state; the second component represented the disordered heads. The ratio of the double integrals of the component spectra was about 50 : 50. It is Fig. 7. Conventional EPR spectra of MSL-fibers in the AdoPP[CH 2 ]P state (A) and in rigor (B). The longer axis of the fibers was oriented parallel to the laboratory magnetic field. Bottom: residual spectrum after digital subtraction (spectrum A ) B). Fig. 6. Conventional EPR spectrum of TCSL-fibers in the AdoPP[CH 2 ]P state. The longer axis of the fibers was oriented parallel to the laboratory magnetic field, except for spectrum B. Spectrum B was recorded on AdoPP[CH 2 ]P-fibers in perpendicular orientation. Digital subtraction of the ADP spectrum (C) from the AdoPP[CH 2 ]P spectrum (A) resulted in a spectrum (D) that was characteristic of randomly oriented spin labels. Bottom: the residual spectrum gener- ated from the spectrum of the ADP–V I state of the fibers (not shown) and spectrum D. The field scan was 10 mT. Ó FEBS 2002 Effect of AdoPP[CH 2 ]P on myosin head domain movements (Eur. J. Biochem. 269) 2173 known that the addition of MgATP plus orthovanadate to rigor solution produced large changes in muscle fibers, and only one spectral component could be detected, which was characteristic of a random population of spin labels [42]. Digital subtraction of the ATP–V i spectrum from the spectrum of TCSL-fibers in the AdoPP[CH 2 ]P state resul- ted in a spectrum similar to that of TCSL-fibers in the ADP state (Fig. 6). Comparison of EPR spectra obtained for homogenized fibers in rigor and the AdoPP[CH 2 ]P state showed a significant decrease in the hyperfine splitting constant 2A¢ zz , which is evidence of the increased rotational mobility. The hyperfine splitting in the AdoPP[CH 2 ]P state was 6.667 ± 0.03 mT (n ¼ 4), whereas a value of 6.780 ± 0.02 mT (n ¼ 6) was estimated in rigor fibers. The apparent rotational correlation time (s r ¼ 0.14 ls) calcu- lated with the Goldman equation [43] corresponds to the detached myosin heads, rotating rapidly, or the binding of AdoPP[CH 2 ]P produces segmental mobility in the environ- ment of the labeled sites. The comparison of spectra recorded in perpendicular orientation of fibers with respect to the laboratory magnetic field in rigor and the AdoPP[CH 2 ]P state (see Figs 5 and 6) showed no appreciable difference. It provides evidence that lineshapes of ST EPR spectra are not affected by the orientation order of probe molecules. The ST EPR spectra of TCSL-fibers in the presence of AdoPP[CH 2 ]P showed changes in the rotational mobility (Fig. 8). The ratio of the low-field diagnostic peaks L¢¢/L changed from 0.800 ± 0.064 (n ¼ 5) to 0.675 ± 0.094 (n ¼ 3). DISCUSSION Characterization of the labeled sites Earlier experiments showed that the maleimide spin labels were located on the reactive cysteine residue (Cys707), and, as a consequence of the modification, the ATPase activities of myosin changed [20,44,45]. TCSL may react with the most reactive lysine residue (Lys84) of myosin: blocking it with TNBS enhanced the Mg 2+ -ATPase activity of myosin by a factor 20 [46]. Our experiments on the Mg 2+ -ATPase activity of TCSL-fibers did not show an increase in the activity in contrast with the findings with untreated fibers. The presence of low concentrations of MgPP i in some way protected the reactive lysine residues from modification [47,48]. The experimental results together with the similar dependence of the probe orientation in MSL-fibers and TCSL-fibers suggested that the paramagnetic labels were probably bound to the same sites. This is supported by the similar value of the ratio of the attached spin labels and that of the decrease in K + /EDTA ATPase activity. The difference in the mean angle of label orientation and the larger angular spread can be explained by the different chemical structure and different attaching linkage of the two labels. Recent results of the effects of Cys707 labeling on myosin showed that, after modification, the myosin heads had limited ability to propel the actin filaments in the in vitro motility assay [19,20,28]. However, the spin-labeling of myosin heads in muscle fibers did not significantly affect the shortening and force generation [11,26]. The lower Ca 2+ ,Mg 2+ -ATPase activity suggests that spin labeling blocked actin activation by preventing the accelerated release of hydrolysis product from the myosin heads. Rotational dynamics of labels in the presence of Ado PP [CH 2 ] P In comparison with rigor fibers, the EPR spectra exhibited significant changes at H || k orientation. Thomas and coworkers came to the conclusion that approximately half of the cross-bridges had a different state [41]. This fraction showed dynamic disorder, the heads being dissociated from actin, whereas the other population had the same spectral feature as in rigor. Our results on AdoPP[CH 2 ]P-fibers led us to a similar conclusion, but the second fraction with a high degree of order exhibited a state similar to the ADP state; the state of these heads differed from that of rigor. AdoPP[CH 2 ]P, similarly to ADP, may induce a change in the orientation of the protein segment that holds the label by means of rotation, resulting in another stable con- formational state. It seems that the nucleotide-induced Fig. 8. ST EPR spectra of spin-labeled fibers. The spectra of TCSL- fibers and MSL-fibers were taken in perpendicular orientations. MgAdoPP[CH 2 ]P was added at a concentration of 16 m M ,andthe incubation of fibers lasted for 10 min before EPR measurement. The decrease in the ratio of the L¢¢/L diagnostic peaks in the AdoPP[CH 2 ]P states shows the increased rotational mobility of spin labels attached to theheadregionofmyosin.Thefieldscanwas10mT. 2174 N. Hartvig et al.(Eur. J. Biochem. 269) Ó FEBS 2002 conformational change is independent of whether ADP or AdoPP[CH 2 ]P binds to the myosin head. However, ADP forms a tight AM–ADP complex, therefore it should favor two-headed binding. The binding of AdoPP[CH 2 ]P is almost complete (85% saturation) at a concentration of 0.5 m M [49,50]. This supports the view that, in the AdoPP[CH 2 ]P state, the myosin heads may have two structural states even in the ordered array of cross-bridges obtained from X-ray analysis [51]. Electron micrographs and X-ray diffraction patterns of insect flight muscle fibers provided evidence that the cross-bridges in the weakly binding state were highly ordered at a uniform 90° angle to the filament axis, but the spin-labeled nucleotide reported disorder. The low-angle X-ray diffraction patterns were modified by AdoPP[CH 2 ]P in insect flight muscle; the ratio of the two inner equatorial peaks was lowered when the concentration of AdoPP[CH 2 ]P was increased [49,52]. Similarly, glycerol-extracted muscle fibers from rabbit psoas muscle gave low-angle X-ray diffraction patterns that differed from the diffraction diagrams of either relaxed or rigor muscle fibers [53]. The results obtained for muscle proteins at lower concentrations of MgAdoPP[CH 2 ]P have recently been criticised [31]. Using X-ray diffraction, the authors conclu- ded that, at saturating concentrations of MgAdoPP[CH 2 ]P (20 m M in the absence of calcium), at high ionic strength and low temperature the cross-bridges bound weakly to actin. This conclusion seems to agree with the EPR results. Barnett & Thomas [42] proposed that a single chemical state of a nucleotide could give rise to more than one conforma- tion of myosin, and a change in the chemical state would alter the equilibrium between these states. Our EPR measurements were performed at two concentrations of MgAdoPP[CH 2 ]P (5 and 16 m M ), but significant differen- ces between the EPR spectra were not observed. The spin labels attached to proteins reflect local conformations and dynamic changes in the microsecond or shorter term regimes. Therefore, the appearance of two populations detected by EPR in the presence of MgAdoPP[CH 2 ]P does not contradict the results of the X-ray diffraction technique. Earlier measurements based on exchangeability of bound nucleotide and 31 P-NMR observations also showed two states of the myosin heads in the absence of nucleotides and ADP or AdoPP[CH 2 ]P states, which implies the rapid interconversion of different conformational states [54,55]. Moreover, the equilibrium, i.e. nearly equal populations at 25 °C between the states, was temperature-dependent: at the lower temperature (4 °C) only one form was detected. Our results are also not in contradiction with earlier EPR results using MSL-fibers. MSL probes reported a more rigid attachment (L¢¢/L ¼ 1.2) to myosin than did TCSL probes (L¢¢/L ¼ 0.83). Therefore, the former may indicate the orientation of the entire head, whereas TCSL may be sensitive to smaller changes in the internal structure. The addition of AdoPP[CH 2 ]P induced increased rotational mobility of both spin labels, which implied alteration of the binding state of myosin to actin. However, it seems that TCSL probes report local change as well, as deduced from the comparison of the rigor and ADP spectra. Precise measurements made on muscle fibers with the use of 15 N and deuterized maleimide spin labels and sophisticated spectral analysis gave evidence that, even in the case of maleimide spin label, small conformational changes were produced on myosin following nucleotide binding to myosin [18,56]. Our data on AdoPP[CH 2 ]P binding suggest that the structural states of the myosin head in rigor, the ADP state and the AdoPP[CH 2 ]P state differ significantly from each other. The differences in the EPR spectra in different states indicate significant alterations in the internal structure of the myosin head region and the properties of binding to actin. Accordingly, our experiments support the view that the rotational motion of myosin reflects functionally relevant conformational changes during ATP hydrolysis by myosin. The TCSL probe attached to the small SH1–SH2 helix may be a sensitive detector that reports the coupling between the motor and regulatory domains of myosin. ACKNOWLEDGEMENTS This work was supported by grants from the National Research Foundation (OTKA T 030248, CO 123) and Ministry of Education (FKFP 0387/2000). REFERENCES 1. Huxley, H.E. (1969) The mechanism of muscle contraction. Science 164, 1356–1365. 2. Huxley, H.E. (1979) Time resolved x-ray diffraction studies on muscle. In Cross-Bridge Mechanism in Muscle Contraction (Sugi, H. & Pollack, G.H., ed.), pp. 391–401. University of Tokyo Press, Tokyo. 3. Rayment, I., Holden, H.M., Whittaker, M., Yohn, C.B., Lorenz, M., Holmes, K.C. & Milligan, R.A. (1993) Structure of the actin- myosin complex and its implications for muscle contraction. Science 261, 58–65. 4. Smith, C.A. & Rayment, I. (1996) X-ray structure of the magne- sium (II) ADP.vanadate complex of the Dictyostelium discoideum myosin motor domain to 1.9 A ˚ resolution. Biochemistry 35, 5404– 5417. 5. Holmes, K.C. (1998) A molecular model for muscle contraction. Acta Crystallogr. A 54, 789–797. 6. Ajtai, K., Peyser, M., Park, S., Burghardt, T.P. & Muhlrad, A. (1999) Trinitrophenylated reactive lysine residue in myosin detects lever arm movement during consecutive steps of ATP hydrolysis. Biochemistry 38, 6428–6440. 7. Bagshaw, C.R. & Trentham, D.R. (1974) The characterization of myosin–product complexes and of product release step during magnesium ion-dependent adenosine triphosphatase reaction. Biochem. J. 141, 331–349. 8. Geeves,M.A.&Holmes,K.C.(1999)Structuralmechanismsof muscle contraction. Annu. Rev. Biochem. 68, 687–728. 9. Roopnarine, O., Szent-Gyo ¨ rgyi, A.G. & Thomas, D.D. (1998) Microsecond rotational dynamics of spin-labeled myosin reg- ulatory light chain induced by relaxation and contraction of scallop muscle. Biochemistry 37, 14428–14436. 10. Fajer, P.G., Fajer, E.A., Schoenberg, M. & Thomas, D.D. (1991) Orientational disorder and motion of weakly attached cross- bridges. Biophys. J. 60, 642–649. 11. Cooke, R. (1986) The mechanism of muscle contraction. CRC Crit. Rev. Biochem. 21, 53–118. 12. Hambly, B., Franks, K. & Cooke, R. (1992) Paramagnetic spin probes attached to a light chain on the myosin head are highly disordered in active muscle fibres. Biophys. J. 63, 1306–1313. 13. Reedy, M.C., Reedy, M.K. & Tregear, R.T. (1988) Two attached non-rigor crossbridge forms in insect flight muscle. J. Mol. Biol. 204, 357–383. 14. Fisher, A.J., Smith, C.A., Thoden, J., Smith, R., Sutoh, K., Holden, H.M. & Rayment, I. (1995) Structural studies of myosin– Ó FEBS 2002 Effect of AdoPP[CH 2 ]P on myosin head domain movements (Eur. J. Biochem. 269) 2175 nucleotide complexes: a revised model for the molecular basis of muscle contraction. Biophys. J. 68, 19s–28s. 15. Ling, N., Shrimpton, C., Sleep, J., Kendrick-Jones, J. & Irving, M. (1996) Fluorescent probes on orientation of myosin regulatory light chains in relaxed, rigor and contracting muscle. Biophys. J. 70, 1836–1846. 16. Holmes, K.C. (1998) A powerful stroke. Nat. Struct. Biol. 5,940– 942. 17. Houdusse, A., Kalabokis, V.N., Himmel, D., Szent-Gyo ¨ rgyi, A.G. & Cohen, C. (1999) Atomic structure of scallop myosin subfragment S1 complexed with MgADP: a novel conformation of the myosin head. Cell 97, 459–470. 18. Burghardt, T.P., Garamszegi, S.P., Park, S. & Ajtai, K. (1998) Tertiary structural changes in the cleft containing the ATP sensi- tive tryptophan and reactive thiol are consistent with pivoting the myosin heavy chain at Gly699. Biochemistry 37, 8035–8047. 19. Bobkov, A.A., Bobkova, E.A., Homsher, E. & Reisler, E. (1997) Activation of regulated actin by SH1-modified myosin subfrag- ment 1. Biochemistry 36, 7733–7738. 20. Bobkova, E.A., Bobkov, A.A., Levitsky, D.I. & Reisler, E. (1999) Effects of SH1 and SH2 modifications on myosin. Similarities and differences. Biophys. J. 76, 1001–1007. 21. Thomas, D.D., Ishiwata, S., Seidel, J.C. & Gergely, J. (1980) Submillisecond rotational dynamics of spin-labeled myosin heads in myofibrils. Biophys. J. 32, 873–889. 22. Thomas, D.D. & Cooke, R. (1980) Orientation of spin-labeled myosin heads in glycerinated muscle fibers. Biophys. J. 32,891– 906. 23. Fajer, P.G. (1994) Determination of spin-label orientation within the myosin head. Proc. Natl Acad. Sci. USA 91, 937–941. 24. Zhao, L., Gollub, J. & Cooke, R. (1996) Orientation of para- magnetic probes attached to gizzard regulatory light chain bound to myosin heads in rabbit skeletal muscle. Biochemistry 35, 10158– 10165. 25. Baker, J.E., Brust-Mascher, I., Ramachandran, S., LaConte, L.E.W. & Thomas, D.D. (1998) A large and distinct rotation of the myosin light chain domain occurs upon muscle contraction. Proc. Natl Acad. Sci. USA 95, 2944–2949. 26. Belagyi, J., Frey, I. & Po ´ to ´ , L. (1994) ADP-induced changes in ordering of spin-labelled myosin heads in muscle fibres. Eur. J. Biochem. 224, 215–222. 27. Mulhern, S.A. & Eisenberg, E. (1978) Interaction of spin-labeled and N-(iodoacetylaminoethyl)-5-naphthylamine-1-sulfonic acid SH1-blocked heavy meromyosin and myosin with actin and ade- nosine triphosphate. Biochemistry 17, 4419–4425. 28. Marriott, G. & Heidecker, M. (1996) Light-directed generation of the actin-activated ATPase. Activity of caged heavy meromyosin. Biochemistry 35, 3170–3174. 29. Trentham, D.R., Bardsley, R.G., Eccleston, J.P. & Weeds, A.G. (1972) Elementary processes of the magnesium ion-dependent adenosine triphosphatase activity of heavy meromyosin. Biochem. J. 126, 635–644. 30. Ma ´ lna ´ si-Csizmadia, A., Wooley, R.J. & Bagshaw, C.R. (2000) Resolution of conformational states of Dictyostelium myosin II motor domain using tryptophan (W501) mutants: implications for the open-closed transition identified by crystallography. Bio- chemistry 39, 16135–16146. 31. Frisbie, S.M., Xu, S., Chalovich, J.M. & YuL.C. (1998) Char- acterization of cross-bridges in the presence of saturating con- centrations of MgAMP-PNP in rabbit permeabilized psoas muscle. Biophys. J. 74, 3072–3082. 32. Zhao, L., Naber, N. & Cooke, R. (1995) Muscle cross-bridges bound to actin are disordered in the presence of 2,3-butanedione monoxime. Biophys. J. 68, 1980–1990. 33. Fajer, P.G. & Marsh, D. (1982) Microwave and modulation field inhomogeneities and effect of cavity Q in saturation transfer EPR spectra. Dependence of sample size. J. Magn. Reson. 49, 212–224. 34. Shimoyama, Y. & Watari, H. (1986) Analysis of saturation transfer electron paramagnetic resonance spectra in term of amplitude and phase. J. Chem. Phys. 84, 3688–3693. 35. Po ´ to ´ ,L.,Frey,I.&Belagyi,J.(1993)ST-EPRwithout phase setting at low power. A calculation method for weak spectra. Abstr. 11th International Biophysics Congress, Budapest. 36. Squier, T.C. & Thomas, D.D. (1986) Methodology for increased precision in saturation transfer electron paramagnetic resonance studies of rotational dynamics. Biophys. J. 49, 921–929. 37. Evans, C.A. (1981) Use of integral of saturation transfer electron paramagnetic spectra to determine molecular rotational correla- tion time. Slowly tumbling spin labels in the presence of rapidly tumbling spin labels. J. Magn. Reson. 44, 109–116. 38. Horva ´ th, L.I. & Marsh, D. (1983) Analysis of multicomponent saturation transfer EPR spectra using the integral method: application to membrane systems. J. Magn. Reson. 54, 363–373. 39. Pollard, T.D., Bhandari, B., Maupin, P., Wachsstock, D., Weeds, A.G. & Zot, H.G. (1993) Direct visualization by electron micro- scopy of the weakly bound intermediates in the actomyosin ade- nosine triphosphatase cycle. Biophys. J. 64, 454–471. 40. Schro ¨ der, R.R., Manstein, D.J., Jahn, W., Holden, H., Rayment, I., Holmes, K.C. & Spudich, J.A. (1993) Three-dimensional atomic model of F-actin decorated with Dictyostelium myosin S1. Nature (London) 364, 171–174. 41. Fajer, P.G., Fajer, E.A., Brunsvold, N.J. & Thomas, D.D. (1988) Effect of AMPPNP on the orientation and rotational dynamics of spin-labeled muscle crossbridges. Biophys. J. 53, 513–524. 42. Barnett, V.A. & Thomas, D.D. (1987) Resolution of conforma- tional states of spin-labeled myosin during steady-state ATP hydrolysis. Biochemistry 26, 314–323. 43. Goldman, S.A., Bruno, G.V. & Freed, J.H. (1972) Estimating slow-motional rotational correlation times for nitroxides by elec- tron spin resonance. J. Phys. Chem. 76, 1858–1860. 44. Sekine, T. & Kielley, W.W. (1964) The enzymic properties of N-ethylmaleimide modified myosin. Biochim. Biophys. Acta 81, 336–345. 45. Crowder, M.S. & Cooke, R. (1984) The effect of myosin sulfhydryl modification on the mechanics of fibre contraction. J. Mus. Res. Cell Motil. 5, 131–146. 46. Fabian, F. & Muhlrad, A. (1968) Effect of trinitrophenylation on myosin ATPase. Biochim. Biophys. Acta 162, 596–603. 47. Miyanishi, T., Inoue, A. & Tonomura, Y. (1979) Differential modification of specific lysine residues in the two kinds of sub- fragment-1 of myosin with 2,4,6-trinitrobenzenesulfonate. J. Bio- chem 85, 747–753. 48. Hozumi, T. & Muhlrad, A. (1981) Reactive lysyl of myosin sub- fragment 1: location on the 27K fragment and labeling properties. Biochemistry 20, 2945–2950. 49. Goody, R.S., Barrington-Leigh, J., Mannherz, H.G., Tregear, R.T. & Rosenbaum, G. (1976) X-ray titration of binding of b,c-imido-ATP to myosin in insect flight muscle. Nature 262, 613–615. 50. Marston, S.B., Rodger, C.D. & Tregear, R.T. (1976) Changes in muscle crossbridges when AMPPNP binds to myosin. J. Mol. Biol. 104, 263–276. 51. Reedy, M.K., Lucaveche, C., Naber, N. & Cooke, R. (1992) Insect crossbridges, relaxed by spin labeled nucleotide, show well- ordered 90° state by X-ray diffraction and electron microscopy, but spectra of electron paramagnetic resonance probes report disorder. J. Mol. Biol. 227, 678–697. 52. Tregear, R.T., Wakabayashi, K., Tanaka, H., Iwamoto, H., Ready, M.C., Ready, M.K., Sugi, H. & Amemiya, Y. (1990) X-ray diffraction and electron microscopy from Lethocerus flight muscle partially relaxed by adenylylimidodiphosphate and ethylene glycol. J. Mol. Biol. 214, 129–141. 2176 N. Hartvig et al.(Eur. J. Biochem. 269) Ó FEBS 2002 53. Lymn, R.W. (1975) Low-angle x-ray diagrams from skeletal muscle: the effect of AMP.PNP, a non-hydrolyzed analogue of ATP. J. Mol. Biol. 99, 567–582. 54. Mihashi, K., Ooi, A. & Hiratsuka, T. (1990) Evidence for the existence of two equilibrium conformations of the ternary complex of myosin subfragment-1, ADP, and orthovanadate. J. Biochem. 107, 464–469. 55. Shriver, J.W. & Sykes, B.D. (1981) Phosphorus-31 nuclear mag- netic resonance evidence for two conformations of myosin sub- fragment-1–nucleotide complexes. Biochemistry 20, 2004–2012. 56. Fajer, P.G., Fajer, E.A., Matta, J.M. & Thomas, D.D. (1990) Effect of ADP on the orientation of spin-labeled myosin heads in muscle fibers: a high resolution study with deuterated spin labels. Biochemistry 29, 5865–5871. Ó FEBS 2002 Effect of AdoPP[CH 2 ]P on myosin head domain movements (Eur. J. Biochem. 269) 2177 . Effect of adenosine 5¢-[b,c-imido]triphosphate on myosin head domain movements Saturation transfer EPR measurements without low-power phase setting No ´ ra. internal dynamics and domain orienta- tion of myosin. In the AdoPP[CH 2 ]P state, approximately half of the myosin heads reflected the motional state of ADP myosin,