Proc. Natl. Acad. Sci. USA
Vol. 94, pp. 9643–9647, September 1997
Biophysics
Independent mobility of catalytic and regulatory domains of
myosin heads
(muscleylight chainyspectroscopyyEPRysaturation transfer–EPR)
BISHOW ADHIKARI*, KALMAN HIDEG†,
AND
PIOTR G. FAJER*‡
*Institute of Molecular Biophysics, Department of Biological Science and National High Magnetic Field Laboratory, Florida State University, Tallahassee, FL
32310; and †Institute of Organic and Medicinal Chemistry, University of Pecs, H-7643, Pecs, Hungary
Communicated by Michael Kasha, Florida State University, Tallahassee, FL, June 20, 1997 (received for review February 14, 1997)
S1-decorated actin filaments and EPR studies have indicated
a rotation of the regulatory domain with respect to the
catalytic domain (9–11). In these cases, the rotation was
observed at the distal end of the regulatory domain on the
binding of ADP only in smooth muscle S1 or brush border S1.
Studies were not carried out in skeletal muscle; thus, the
question remained whether there is a similar rotation in
skeletal muscle and on the location of the possible hinge. The
atomic structure of myosin S1 strongly suggests the interface
between the catalytic domain and the essential light chain is a
putative hinge (12).
We spin labeled two cysteine residues, 31Å apart, on either
side of the putative hinge—Cys-707 in the catalytic domain and
Cys-177 of the essential light chain in the regulatory domain—
and compared their mobilities in filaments made of skeletal
myosin. Saturation transfer–EPR (ST-EPR) spectroscopy was
utilized in probing domain movements because ST-EPR is
uniquely sensitive to the microsecond motions expected for the
myosin head. We found that the mobility of the catalytic
domain is three to five times higher than that of the regulatory
domain. This finding implies that the connection between the
two domains is flexible, lending further support to models in
which internal shape changes within the myosin head lead to
force generation (4, 13, 14).
ABSTRACT
The recent determination of the myosin head
atomic structure has led to a new model of muscle contraction,
according to which mechanical torque is generated in the
catalytic domain and amplified by the lever arm made of the
regulatory domain [Fisher, A. J., Smith, C. A., Thoden, J.,
Smith, R., Sutoh, K., Holden, H. M. & Rayment, I. (1995)
Biochemistry 34, 8960–8972]. A crucial aspect of this model is
the ability of the regulatory domain to move independently of
the catalytic domain. Saturation transfer–EPR measurements
of mobility of these two domains in myosin filaments give
strong support for this notion. The catalytic domain of the
myosin head was labeled at Cys-707 with indane dione spin
label; the regulatory domain was labeled at the single cysteine
residue of the essential light chain and exchanged into myosin.
The mobility of the regulatory domain in myosin filaments was
characterized by an effective rotational correlation time (tR)
between 24 and 48 ms. In contrast, the mobility of the catalytic
domain was found to be tR 5 5–9 ms. This difference in
mobility between the two domains existed only in the filament
form of myosin. In the monomeric form, or when bound to
actin, the mobility of the two domains in myosin was indistinguishable, with tR 5 1–4 ms and >1,000 ms, respectively.
Therefore, the observed difference in filaments cannot be
ascribed to differences in local conformations of the spinlabeled sites. The most straightforward interpretation suggests a f lexible hinge between the two domains, which would
have to stiffen before force could be generated.
METHODS
Protein Purification. All preparative procedures were carried out at 4°C. Rabbit myosin was extracted with Guba-Straub
solution (0.3 M KCl, 0.05 M K2HPO4, 0.1 M KH2PO4) and
purified by repeated polymerization–depolymerization cycles
in low and high ionic strength (m) buffers as described by
Margossian and Lowey (15). Purified myosin was either directly used or stored in 50% glycerol at 220°C. A mixture of
essential light chains (ELCs) containing A1 and A2 isoforms
was prepared from purified myosin by removal of myosin heavy
chain from the light chains by guanidine-HCl denaturation
followed by ethanol precipitation (16), except for an additional
repeat of the procedure for the regulatory light chain removal.
The procedure yielded an '95% pure essential light chain
mixture upon examination by 12% SDSyPAGE gels. Purified
ELC (1–2 mgyml) was stored at 270°C in 5 mM phosphate
buffer, 1 mM NaN3, and 0.1 mM DTT at pH 7.0 (storage
buffer). Myosin rods were prepared by chymotryptic digestion
of myosin to remove S1 and soluble heavy meromyosin followed by removal of residual myosin by solubilization and
denaturation in high salt and 95% ethanol (15). Actin was
The myosin head (subfragment 1, S1) plays a central role
during muscle contraction. It is thought that during the
ATPase cycle S1 interacts with actin and undergoes a series of
specific changes in its conformation, resulting in a directional
strain on the actin filaments. The nature of these conformational changes and the mechanism by which they produce the
directed force is still unclear (1, 2), although the original model
of Huxley and Simmons (3) provides a valid framework. The
determination of the atomic structures of S1, actin, and the
low-resolution structure of the acto-S1 complex has led to a
hypothesis of a structural model, in which helix movement is
generated by the hydrolysis of ATP and the formation of a
stereospecific acto-S1 complex generates torque within the
catalytic domain, which is then amplified by the regulatory
domain acting as a lever arm (4, 5).
Rayment’s model was preceded by the observation of shape
changes in the myosin heads between various chemical states
of myosin. A decrease in the hydrodynamic size of S1 upon
MgADP binding and upon ATP hydrolysis was observed by
Highsmith and Eden (6) and by Wakabayashi et al. (7, 8). More
recently, low-resolution electron microscopy reconstitution of
Abbreviations: DITC, diisothiocyanate; MSL, N-(1-oxy-2,2,5,5tetramethyl-4-pyperidinyl)maleimide; IASL, N-(1-oxy-2,2,6,6tetramethyl-4-piperidinyl)iodoacetamide; InVSL, 2-(-oxyl-2,2,5,5tetramethyl-3-pyrrolin-3-methynyl)indane-1,3-dione; ST-EPR, saturation transfer–EPR; S1, myosin subfragment 1; ELC, essential light
chain.
‡To whom reprint requests should be addressed. e-mail: fajer@
magnet.fsu.edu.
The publication costs of this article were defrayed in part by page charge
payment. This article must therefore be hereby marked ‘‘advertisement’’ in
accordance with 18 U.S.C. §1734 solely to indicate this fact.
© 1997 by The National Academy of Sciences 0027-8424y97y949643-5$2.00y0
PNAS is available online at http:yywww.pnas.org.
9643
9644
Biophysics: Adhikari et al.
Proc. Natl. Acad. Sci. USA 94 (1997)
prepared from acetone powder by extracting in a low m buffer
followed by cycles of polymerization and depolymerization
induced by m changes according to Pardee and Spudich (17).
Spin Labeling. Purified ELC mixture in storage buffer was
dialyzed in 40 mM KCl, 0.2 mM EDTA, 1 mM NaN3, and 5 mM
phosphate buffer at pH 7.0 and labeled with a 5-fold molar
excess of spin label for 5–12 hr, followed by dialysis in the same
buffer. A labeling ratio of 80–100% was achieved, as determined by spin quantitation using EPR and protein concentration determination using the BCA protein assay (Pierce). The
catalytic domain of S1 was labeled with indane dione spin label
(InVSL) according to Roopnarine et al. (18). Myosin rods were
labeled in 0.6 M KCl, 50 mM Mops, pH 7.0, with a 5-fold excess
of InVSL overnight, followed by dialysis in the same buffer to
remove the excess spin label. To form filaments, labeled
myosin and rods were dialyzed in 40 mM KCl, 10 mM Mops,
pH 7.0, centrifuged at 10,000 3 g for 10 min., and loaded into
50-ml fused silica capillary tubes (Wilmad, Buena, NY) for
EPR experiments.
Spin-labeled ELCs (InVSL-ELCs) were immobilized on
preactivated diisothiocyanate (DITC) glass beads (Sigma) by
incubation of the InVSL-ELCs in buffer-prewashed DITC
beads overnight, at an equimolar concentration of protein and
binding sites on the beads, followed by removal of free light
chains by several washes in the same buffer.
The spin-labeled light chains were exchanged into myosin
according to the method of Wagner (16). Typically, 40–80%
exchange of labeled ELC per head was achieved, as determined by spin quantitation and the BCA assay. The exchange
of labeled ELC into myosin did not significantly (,10%) alter
the ATPase activity (Table 1).
EPR Experiments. EPR and ST-EPR experiments were
carried out as in Adhikari and Fajer (19), using a TE102 cavity
with microwave field strength (H1) of 0.03–0.08 G and microwave modulation amplitude (Hm) of 2 G for EPR and H1 5
0.25 G and Hm 5 5 G for ST-EPR. The modulation frequency
was 50 kHz. All experiments were carried out at 4°C. The
ST-EPR spectra were analyzed by comparison of the experimental line shapes with those obtained from InVSL-labeled
hemoglobin tumbling in media of known viscosity [87.55%
wtywt, glycerolywater mixture, 230° to '20°C (20)]. Conventional EPR spectra were analyzed as described by Freed (21),
with rotational correlation time given by tR 5 a(1 2 2Teffy
2Tmax)b, where 2Tmax is the rigid limit for a particular spinlabeled protein and 2Teff is the splitting between the outermost
peaks, a 5 5.4 3 10210 s, and b 5 21.36 for Brownian motion.
RESULTS
Spin-Labeling ELCs. Three spin labels, InVSL, iodoacetemide (IASL), and maleimide (MSL) nitroxides, were tested
in an attempt to find a rigidly attached probe on ELCs that will
accurately report domain or global motions rather than the
Table 1. ATPase activities* of InVSL-ELC-exchanged myosin and
the native myosin
Condition
Native
InVSL-ELC
exchanged
Physiological ionic strength†
Low ionic strength†
Actomyosin in low ionic
strength
0.016 6 0.002
0.026 6 0.006
0.016 6 0.002
0.022 6 0.006
0.5 6 0.03
0.4 6 0.02
*ATPase activities are reported as mmol of Piymin per mg of myosin.
The actomyosin activity was carried out with a 7.5–10 molar excess of
actin over myosin. All experiments were carried out at 25°C.
†The physiological ionic strength buffer consisted of 130 mM potassium propionate, 20 mM Mops, 2 mM MgCl2, 2 mM EGTA, 1 mM
NaN3, pH 7.0. The low-ionic-strength buffer was 40 mM salt and 10
mM Mops, pH 7.0.
FIG. 1. Rigidity of InVSL attached to Cys-177 of LC1. Conventional EPR spectra of InVSL-labeled ELCs in solution (A) and after
exchange into myosin filaments formed by dialysis (B). The narrow
effective splitting (2Teff) in the solution spectrum, which indicates
rapid nanosecond motions, becomes broad and large in the myosin
filament spectrum, thus indicating the absence of fast nanosecond
motions. Immobilization of the labeled ELCs on DITC-coated glass
beads also resulted in an EPR spectrum (C) with a large 2Teff,
indicating rigid attachment of the label. Likewise, the ST-EPR spectrum (D) from the immobilized sample with the L0yL line-height ratio
of 1.5 is characteristic of an effective rotational correlation time (tR)
in the millisecond time scale. The samples were in 40 mM KCl, 10 mM
Mops, pH 7.0, and 23°C.
independent motion of the spin label relative to the protein. In
solution, the spectra of labeled ELCs indicated rapid nanosecond motions in the following order of increasing rotational
correlation times: InVSL-ELC, 1.2 ns; IASL-ELC, 1.3 ns; and
MSL-ELC, 1.5 ns (Fig. 1A, IASL and MSL spectra not shown).
These mobilities are faster than the tR 5 8 ns predicted from
the Stokes–Einstein equation.§ Faster than expected mobilities
of spin-labeled protein imply librational motion with respect to
the isolated light chain or very fast mobility of the spin-labeled
domain.
Fortunately, this fast mobility was reduced when the InVSLELC was exchanged into myosin (Fig. 1B) or into S1 (spectra
not shown). The largest immobilization was observed with
InVSL. Compared with the sharp spectrum in solution, the
spectrum from myosin filaments was broad, with the hyperfine
splitting, 2Teff 5 71.0 6 0.21 G, which is the rigid limit for this
label. The spectra of IASL and MSL were only slightly
broadened (not shown), implying considerable librational motions relative to the protein, tR 5 1.9 ns and tR 5 1.6 ns,
respectively, rendering them unsuitable for the study of regulatory domain dynamics.
To further examine the rigidity of InVSL with respect to
ELC, the spin-labeled protein was immobilized on DITCcoated glass beads. Both the conventional and ST-EPR spectra
were at the rigid limit (Fig. 1 C and D, respectively). The
splitting of the EPR spectrum is 70.7 6 0.22 G and the
diagnostic line height ratio, L0yL, of the ST-EPR spectrum is
1.5. Thus, the InVSL label is immobilized both on the nanosecond time scale of EPR and on the millisecond time scale of
ST-EPR experiments, and therefore can be used to study the
dynamics of the regulatory domain. The InVSL was previously
shown to be rigidly attached to Cys-707 of the catalytic domain
using the same criteria (18).
§tR 5 M(v 1 h)hykT, where h denotes the viscosity of the medium; M,
the mass of ELC (19, 3 kDa); v, the partial specific volume (0.74
cm3yg); h, the hydration (0.2 g wateryg protein); k, the Boltzmann
constant; and T, the absolute temperature.
Biophysics: Adhikari et al.
Proc. Natl. Acad. Sci. USA 94 (1997)
9645
FIG. 3. ST-EPR spectra of the regulatory (Upper) and the catalytic
domains (Lower) of myosin heads in monomeric myosin. The samples
were in 0.6 M KCl, 20 mM Mops, pH 7.0, and 4°C.
FIG. 2. ST-EPR spectra comparing the mobility of the regulatory
domain (Top), the catalytic domain (Middle), and the rod (Bottom) in
synthetic filaments prepared by dialysis at m 5 45 mM (see Methods).
The spectrum of the regulatory domain indicates lower mobility for
this domain than that of the catalytic domain. The samples were in 40
mM KCl, 10 mM Mops, pH 7.0, and 4°C.
Mobility in Myosin Filaments. The spectra of the myosin
filaments formed at low ionic strength (m 5 45 mM) with spin
labels in the catalytic domain (Cys-707), regulatory domain
(Cys-177), and myosin rod are shown in Fig. 2. A significant
difference in mobility is apparent between the spectra of the
catalytic and regulatory domains. The higher intensity of the
diagnostic feature L0 in the spectrum of the regulatory domain
reflects the slower mobility of this domain as compared with
the catalytic domain spectrum. The effective tR, calculated
using the L0yL line height ratio, is 36 6 12 ms for the regulatory
domain and 6.9 6 1.5 ms for the catalytic domain (Table 2).
The latter value is in agreement with a previous ST-EPR
estimate (22) or from time-resolved optical anisotropy (23, 24).
The mobility of both domains of the myosin head was higher
than the mobility of myosin rods, tR 5 165 6 50 ms, consistent
with a hinge at the head-rod junction, enabling myosin head
movement relative to the myosin filament backbone (Fig. 2).
The flexibility of that particular hinge is modulated by ionic
strength, pH, phosphorylation of regulatory light chain, or
presence of MgATP (25).
Mobility in Myosin Monomers and Actomyosin. It is possible that the difference in mobility between the two domains
arises from the interaction of the label with its immediate
environment of various amino acid residues surrounding the
labeled sites. For example, the different orientation of the spin
label principal axes with respect to the diffusion axis might
result in a different ST-EPR line shape for the same motional
mode (26). To exclude this possibility, the mobility of the two
domains was measured in monomeric myosin and in actomyosin. If the line-shape difference was due to the local environment, then we would expect the difference line shape to
persist in actomyosin or monomeric myosin. Fig. 3 shows a
representative set of spectra for the two domains in monomeric
myosin.
Both spectra indicate faster mobility than was observed for
the filament form of myosin, consistent with greater rotational
flexibility in the monomeric form than in the filament form.
Importantly, no significant difference was observed in the
spectra between the two domains. The correlation times were
2.6 6 0.6 ms for the regulatory domain and 2.7 6 1.6 ms for the
catalytic domain (Table 2). Likewise, when the myosin head is
immobilized by actin, the mobility of the two domains is equal;
tr 5 1–2 ms (Fig. 4).
Table 2. Comparison of effective correlation times of the two
head domains
Effective tR, ms
Condition
Regulatory
domain
Catalytic
domain
Myosin monomers*
Myosin filaments†
Actomyosin*
2.6 6 0.6 (3)
36 6 12 (6)
1,000–2,000 (3)
2.7 6 1.6 (3)
6.9 6 1.5 (14)
1,000–3,000 (2)
Effective correlation time is defined as that of InVSL-hemoglobin
displaying the same L0yL line-height ratio. The numbers enclosed in
brackets indicate the number of measurements.
*In 0.6 M KCl, 20 mM Mops, pH 7.0 at 4°C.
†In 40 mM KCl, 10 mM Mops, pH 7.0 at 4°C.
FIG. 4. ST-EPR spectra of the regulatory (Upper) and the catalytic
(Lower) domains of myosin in the presence of actin filaments. The
samples were in 0.6 M KCl, 20 mM Mops, pH 7.0, and 4°C.
9646
Biophysics: Adhikari et al.
The absence of any mobility differences between the catalytic and regulatory domains in the two motional extremes—
monomeric form and actomyosin—strongly suggests that neither differences in the local environment of the label sites nor
the different anisotropy of label motion at the two sites
contributes toward the difference observed in the myosin
filaments (26).
DISCUSSION
The present study demonstrates that the regulatory and catalytic domains of the myosin head are capable of independent
movement, indicating that the connection between the two
domains is flexible. The difference in the measured effective
tR from the two domains may be due to differences in rates of
motion or amplitude, or a combination of both (27). The
possibility of changing probe environment to account for this
difference was ruled out by identical tR in the two domains in
actomyosin and monomeric myosin, suggesting that the local
environment of the probed sites in catalytic and regulatory
domains are the same. The finding that tR of the two domains
varies as expected in monomers, in filaments, and when bound
to actin, strengthens the validity of using rotational correlation
time for probing the relative movement of myosin head
domains in various conformations.
The mobility of the two domains might be modeled using the
formalism developed by Broersma (28) as modified by Hagerman and Zimm (29). To model the relative motions of the
regulatory and catalytic domains, let us assume that the
mobility of the regulatory domain equals that of the entire
head, which is approximated by a semi-rigid cylinder with 5-nm
diameter and 17-nm length and is attached to the thick
filament shaft by a universal joint (S1yS2 junction). The
catalytic domain might be modeled as a rigid cylinder half the
length of the head (8.5 nm) and attached to the regulatory
domain by a second universal joint (catalytic-regulatorydomains hinge). The predicted difference in the diffusion
coefficients between the regulatory and catalytic domains is
then 5.5-fold, while the observed values differ by 5.3. This good
agreement between the observed and predicted values supports the validity of the original assumption of low motional
restrictions placed on the mobility of the catalytic domain by
the regulatory domain.
Myosin heads in filaments possess considerable freedom of
movement even though the rod and part of S2 form the rigid
filament backbone, as established previously for the catalytic
domain (22–24) and in this work for the regulatory domain.
The larger increase in tR of the regulatory domain compared
with the catalytic domain upon filament formation indicates
that the hinge at S1yS2 may not be completely flexible,
perhaps due to interactions between this domain and the
filament backbone. Indeed, phosphorylation of regulatory
light chain, which results in the movement of the head away
from the filament backbone and an increase of regulatory
domain mobility, suggests that these interactions may be
related to force modulation (25, 30).
The most recent model of muscle contraction, based on the
crystal structure of the myosin head, postulates a domain
displacement as necessary for the generation of the power
stroke (4). A comparison of the crystal structures of the
catalytic domain from Dictostelium discoideum complexed
with beryllium fluoride and aluminum fluoride, thought to
mimic the ATP bound state and the transition state for
hydrolysis, respectively, indicates significant movement of the
lower 50-kDa domain. This movement is accompanied by
bending at Ile-464 and Gly-466 (chicken S1 sequence numbering), resulting in a rotation ('20°) and a translation (2.5 Å)
of the helix between SH1 and SH2. Although the crystal
structure was missing the regulatory domain, extrapolation to
the position of the long helix of the heavy chain constituting the
Proc. Natl. Acad. Sci. USA 94 (1997)
regulatory domain suggests that such a movement might alter
the position of the regulatory domain relative to the motor
domain (5, 31). In a number of studies of isolated myosin
heads, significant changes in the hydrodynamic radius were
observed (6–8, 32). Such changes have been postulated to
generate force (6, 14). Most recently, the displacement of the
regulatory domain with respect to the catalytic domain induced by MgADP has been demonstrated in smooth S1
decorating F-actin by three-dimensional electron microscopy
reconstruction (9). Similar observations have also been made
in skinned fibers with spin-labeled LC2 exchanged into smooth
S1 (11). Importantly, domain rearrangement in skeletal muscle
was not observed; thus, the reported movement in smooth
muscle cannot be the universal conformational change responsible for force generation.
As concluded by both of the above groups, it is not clear
whether the observed regulatory movement is responsible for
force generation. Our own results do not correlate the structural changes with force generation, because the observed
mobility differences between the domains existed in the absence of actin and ATP. However, the presence of a flexible
hinge within the myosin head suggests one of the proposed
conformational changes during the power stroke must increase
the stiffness between the two domains. We might speculate
that the hinge stiffness is a function of the attachment to actin
filaments andyor the state of ATP hydrolysis products in the
active site of myosin as recently implied by Highsmith and
Eden (33). The absence of differences between the domains in
acto-myosin complexes, implying rigidification of the hinge,
hints at this scenario. The possibility of the modulation of this
hinge by the intermediate states of the ATPase cycle—in the
presence of ADP, aluminofluoride.ATP, and Va.ATP—are
being explored. The functional significance of the hinge is not
clear, but the observed flexibility might assist in the formation
of a stereospecific acto-myosin interface.
Reorientations of the regulatory domain in fibers during
contraction has been demonstrated by a number of studies
utilizing EPR and time-resolved fluorescence (34–37). These
and the finding of direct proportionality between force and
length of the regulatory domain (38) suggest that the movement of the regulatory domain is responsible for the power
stroke. The capability of relative movement indicates that the
regulatory domain can undergo power-stroke transition while
the catalytic domain is attached to actin.
In summary, the present ST-EPR study demonstrates that in
skeletal myosin filaments the regulatory and catalytic domains
of myosin exhibit unequal mobility; the regulatory domain
moves slower than the catalytic domain. These results are
interpreted as a strong indication that the two domains of the
myosin head are connected by a flexible hinge.
We thank Drs. Brett Hambly and Ian Trayer for critical reading of
the manuscript and many discussions; Dr. Stefan Highsmith for sharing
his data prior to publication; Dr. Danuta Szczesna for help with initial
experiments; and Liz Fajer and April Adhikari for help in editing the
manuscript. This research was sponsored by the National Science
Foundation (NSF-IBN-9507477) (to P.F.), the American Heart Association (GIA-9501335) (to P.F.), a student fellowship from the American Heart Association (to B.A.), and the Hungarian National Research Foundation OTKA T-017842 (to K.H).
1.
2.
3.
4.
5.
6.
Cooke, R. (1993) Curr. Opin. Cell Biol. 3, 590–592.
Howard, J. & Spudich, J. A. (1996) Proc. Natl. Acad. Sci. USA 93,
4462–4464.
Huxley, A. F. & Simmons, R. M. (1971) Nature (London) 233,
533–538.
Rayment, I., Holden, H. M., Whittaker, M., Yohn, C. B., Lorenz,
M., Holmes, K. C. & Milligan, R. A. (1993b) Science 261, 58–64.
Fisher, A. J., Smith, C. A., Thoden, J., Smith, R., Sutoh, K.,
Holden, H. M. & Rayment, I. (1995) Biophys. J. 68, 19s–28s.
Highsmith, S. & Eden, D. (1990) Biochemistry 29, 4087–4093.
Biophysics: Adhikari et al.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
Wakabayashi, K., Tokunaga, M., Kohno, I., Sugimoto, Y., Hamanaka, T., Takezawa, Y., Wakabayashi, T. & Amemiya, Y.
(1992) Science 258, 443–447.
Sugimoto, Y., Tokunaga, M., Takezawa, Y., Ikebe, M. & Wakabayashi, K. (1995) Biophys. J. 68, 29s–34s.
Jontes, J. D., Wilson-Kubalek, E. M. & Milligan, R. A. (1995)
Nature (London) 378, 751–753.
Whittaker, M., Wilson-Kubalek, E. M., Smith, J. E., Faust, L.,
Milligan, R. A. & Sweeney, H. L. (1995) Nature (London) 378,
748–751.
Gollub, J., Cremo, C. R. & Cooke, R. (1996) Nat. Struct. Biol. 3,
796–802.
Rayment, I., Rypniewski, W. R., Schmidt-Base, K., Smith, R.,
Tomchick, D. R, Benning, M. M., Winkelmann, D. A., Wesenberg, G. & Holden, H. M. (1993) Science 261, 50–58.
Highsmith, S. & Cooke, R. (1983) Cell Muscle Motil. 4, 207–237.
Cooke, R. (1986) CRC Crit. Rev. Biochem. 21, 53–118.
Margossian, S. & Lowey, S. (1982) Methods Enzymol. 85, 55–71.
Wagner, P. D. (1982) Methods Enzymol. 85, 72–81.
Pardee, J. D. & Spudich, J. A. (1982) Methods Enzymol. 85,
164–181.
Roopnarine, O., Hideg, K. & Thomas, D. D. (1993) Biophys. J. 64,
1896–1907.
Adhikari, B. B. & Fajer, P. G. (1996) Biophys. J. 70, 1872–1880.
Fajer, P. G. & Marsh, D. (1983) J. Magn. Reson. 55, 205–215.
Freed, J. (1976) in Spin Labeling: Theory and Applications, ed.
Berliner, L. (Academic, New York), Vol. 1, pp. 53–132.
Thomas, D. D., Seidel, J. C., Gergely, J.,& Hyde, J. S. (1975) J.
Supramol. Struct. 3, 376–390.
Highsmith, S., Kretzschmar, K. M., O’Konski, C. T. & Morales,
M. F. (1977) Proc. Natl. Acad. Sci. USA 74, 4986–4990.
Proc. Natl. Acad. Sci. USA 94 (1997)
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
9647
Eads, T. M., Thomas, D. D. & Austin, R. H. (1984) J. Mol. Biol.
179, 55–81.
Adhikari, B., Somerset, J., Hideg, K. & Fajer, P. G. (1996)
Biophys. J. 70, A265 (abstr.).
Robinson, B. H., Thomann, H., Beth, A. H., Fajer, P. G. &
Dalton, L. R. (1985) in EPR Studies of Biological Systems (CRC,
Boca-Raton, FL).
Howard, E. C., Lindahl, K. M., Polnaszek, C. F. & Thomas, D. D.
(1993) Biophys. J. 64, 581–593.
Broersma, S. (1960) J. Chem. Phys. 32, 1632–1635.
Hagerman, P. J. & Zimm, B. H., (1981) Biopolymers 15, 1481–
1502.
Levine, R. J. C., Kensler, R. W., Yang, Z., Stull, J. T. & Sweeney,
H. L. (1996) Biophys. J. 71, 898–907.
Fisher, A. J., Smith, C. A., Thoden, J. B., Smith, R., Sutoh, K.,
Holden, H. M. & Rayment, I. (1995) Biochemistry 34, 8960–8972.
Suda, H., Sugimoto, M., Chiba, M. & Uemura, C. (1995) Biochem. Biophys. Res. Commun. 211, 219–225.
Eden, D. & Highsmith, S. (1997) Biophys. J. 73, 952–958.
Hambly B., Franks K. & Cooke R. (1992) Biophys. J. 63,
1306–1313.
Irving, M., Allen, T. S., Sabido-David, C., Craik, J. S., Brandmeier, B., Kendrick-Jones, J., Corrie, J. E. T., Trentham, D. R.
& Goldman, Y. E. (1995) Nature (London) 375, 688–691.
Ling, N., Shrimpton, C., Sleep, J., Kendrick-Jones, J. & Irving, M.
(1996) Biophys. J. 70, 1836–1846.
Allen, T. S., Ling, N., Irving, M. & Goldman, Y. E. (1996)
Biophys. J. 70, 1847–1862.
Uyeda, T. Q. P., Abramson, P. D. & Spudich, J. A. (1996) Proc.
Natl. Acad. Sci. USA 93, 4459–4464.