3190 Biophysical Journal Volume 84 May 2003 31903196

EPR Spectroscopy Shows a Microtubule-Dependent Conformational

Change in the Kinesin Switch 1 Domain

Nariman Naber,* Sarah Rice,y Marija Matuska,* Ronald D. Vale,y Roger Cooke,*z and Edward Pate
*Department of Biochemistry and Biophysics, yDepartment of Cellular and Molecular Pharmacology, and zCardiovascular Research
Institute, University of California, San Francisco, California 94143; and Department of Mathematics, Washington State University,
Pullman, Washington 99164

ABSTRACT We have used site-directed spin-labeling and electron paramagnetic resonance spectroscopy to monitor
a conformational change at the nucleotide site of kinesin. Cys-lite kinesin (K349 monomer) with the mutation S188C was spin
labeled with MSL or MTSL. This residue is at the junction between the switch 1 region (which is a structure known to be
sensitive to bound nucleotide in the G-proteins) and the a3-helix, adjacent to the nucleotide site. The spectra showed two or
more components of mobility, which were independent of nucleotide in the absence of microtubules (MTs). The spectra of both
labels showed a change of mobility upon binding to MTs. A more mobile spectral component became enhanced for all
triphosphate analogs examined, AMPPNP, ADP AlFx, or ADP BeFx, in the presence of MTs, although the magnitude of the
d d

new component and the degree of mobility varied with nucleotide analog. The ADP state showed a much-reduced spectral
change with a small shift to the more immobilized component in the presence of MTs. For kinesin ADP MT, a vant Hoff plot
d d

gave DH 8
96 kJ/mol implying that the conformational change was extensive. We conclude there is a conformational change
in the switch 1-a3-helix domain when kinesin binds to MTs.

INTRODUCTION
One fundamental goal in the field of motility is to define the adjacent to, and displaced from, the nucleotide triphosphate
conformational changes at the nucleotide site associated with binding location.
the interaction of the motor with the physiological substrate, Thus, structural comparisons between myosin, the G-pro-
and how the conformational changes that accompany nu- teins, and kinesin-family motors suggest the possibility of
cleotide hydrolysis are translated into movement. Kinesin- considerable domain flexibility in the switch 1 region of
family motors, along with myosin motors and the guanosine kinesin-family motors. Indeed, a disordered switch 1 region
triphosphate binding proteins, are members of the larger has recently been seen in the crystal structure of Kar3 con-
G-protein superfamily (Vale, 1996; Kull et al., 1998). In this taining the additional mutation of a universally conserved
superfamily, the triphosphate-binding domain is composed of arginine in switch 1 to an alanine (Yun et al., 2001). Addition-
three conserved amino acid sequences: the P-loop, switch 1, ally, we have previously used molecular dynamics to suggest
and switch 2. A comparison of crystal structures shows that that the switch 1 region of the kinesin motors could adopt a
the conformation of switch 1 can vary. In the initial structures closed conformation adjacent to the nucleotide triphosphates
of kinesin-family motors, the phosphate-binding portion of as seen in the structures of myosin and in some G-proteins,
the nucleotide site formed an open trough, with the and that such a conformation was thermodynamically stable
triphosphates exposed to solvent (Gulick et al., 1998; Kull (Minehardt et al., 2001). Electron paramagnetic resonance
et al., 1996; Sablin et al., 1996; Sack et al., 1997) and the spectroscopy (EPR) studies of the changes in the mobility of
switch 1 region (amino acids 190204 in human kinesin) nucleotide-analog, and EPR probes bound at the active site of
displaced away from the nucleotide. In all myosin motor kinesin-family motors (Naber et al., 2002), have reported
x-ray structures, the switch 1 region is found to be adjacent to a conformational change at the active site upon binding to
the nucleotide, forming in conjunction with the P-loop and microtubules (MTs). Switch 1 was not directly monitored in
switch 2 a closed phosphate tube (Yount et al., 1995) in which the studies, but the conformational change observed was
the phosphates are tightly bound. In G-protein x-ray compatible with the closed switch 1 conformation proposed
structures, switch 1 is sensitive to both the bound nucleotide
by Minehardt et al. (2001). This fully closed structure has
(Sprang, 1997) and nucleotide exchange factors (Goldberg,
been proposed to be essential for both nucleotide hydrolysis
1998; Boriack-Sjodin et al., 1998), and is located both
and for the correct coordination of the conserved glycine in
switch 2 to the g-phosphate of the nucleotide. By analogy
with nucleotide-site conformational changes in both myosin
Submitted October 21, 2002, and accepted for publication January 15, and the G-proteins, a popular hypothesis is that this latter
2003.
coordination is an essential element for motility (Vale and
Address reprint requests to Nariman Naber, Dept. of Biochemistry and
Biophysics, Campus Box 2240, University of California, San Francisco,
Milligan, 2000; Smith and Rayment, 1996; Kull and Endow,
CA 94143. Tel.: 415-476-1975; Fax: 415-476-1902; E-mail: naber@itsa. 2002; Woehlke, 2001; Holmes and Geeves, 2000; Cooke,
ucsf.edu. 1997; Schief and Howard, 2001).
2003 by the Biophysical Society EPR spectroscopy has been a useful tool in monitoring
0006-3495/03/05/3190/07 $2.00 conformational changes in motor proteins (Baumann et al.,
Kinesin Switch 1 Conformational Change 3191

2001; Cooke et al., 1982; Fajer, 1994; Baker et al., 1998;

Naber et al., 1997; Gollub et al., 1999; Thomas et al., 1983;
Berger and Thomas, 1994). In the present studies we use
EPR spectroscopy as a probe of switch 1 conformational
changes in kinesin. An EPR spin probe can be introduced
onto the motor protein by covalent attachment to a cysteine
residue in the protein. The motion of the attached probe is
restricted by the adjacent protein surface. Conformational
changes can thus be detected via changes in probe mobility,
monitored as changes in the EPR spectrum.
Covalent attachment of an EPR probe at an endogenous
cysteine suffers from two major deficiencies. First, the cys-
teine may not be at the precise location in the protein that one
wishes to investigate. Second, the presence of multiple
cysteines can result in multiple labels per protein, each with
their own distinct spectral signature, dramatically complicat-
FIGURE 1 Ribbon diagram of S188C kinesin. Human kinesin (Kull et al.,
ing interpretation of spectral changes. Thus the use of endo- 1996) with Cys-188 (yellow) at the junction of the a3-helix (cyan, aa 176
genous residues has been augmented by site-directed spin 189) with the switch 1 region (magenta, aa 190213). The side chain of
labeling (SDSL) approaches (Serag et al., 2002; Isas et al., Ser-188 projects out into the aqueous environment in the crystal structure.
2002; Hubbell et al., 1996). Endogenous, reactive cysteines The nucleotide-binding domain contains the P-loop (blue), switch 1, and
switch 2 (green) homologous to myosin-family motors and the G-proteins.
are replaced with nonreactive amino acids (Cys-lite pro- The b5-ribbon, aa 205216, is shown in orange.
teins (Rice et al., 1999)), and a single cysteine is then intro-
duced at a specific desired location. The presence of a single
reactive element simplifies spectral analysis and allows for METHODS
more aggressive conditions for covalent labeling, enhancing The mutation S188C was introduced into monomeric Cys-lite K349 kine-
EPR signal strength. SDSL has proved to be particularly pow- sin, and the protein was bacterially expressed and purified as previously
erful when x-ray crystal structures are available, and suggest described (Rice et al., 1999); amino acids 1349. Kinesin was initially
not only a region involved in a potential conformational dialyzed into buffer A containing 25 mM PIPES, 100 mM NaCl, 2 mM
MgCl2, 1 mM EGTA, 50 mM ADP, pH 7.0, 48C. The protein was spin
change, but also allowing structural fine-tuning in the specific
labeled by reacting overnight at 48C after addition of a twofold molar excess
choice of amino acid to be mutated. We take advantage of of 4-maleimido-2, 2, 6, 6-tetramethyl-1-piperidinyloxy (MSL) or 1-oxyl-2,
these factors in the present work. 2, 5, 5-tetramethyl-D3-pyrroline-3-methyl methane thiosulfonate (MTSL).
In kinesin, the switch 1 region is located between Repetitive steps of protein concentration via centrifugation through a 10 kDa
a b-ribbon (b5) that forms part of the central, core b-sheet cutoff sizing filter, and subsequent dilution with buffer A, were used to
remove excess spin label. The final protein concentration was ;100 mM.
of the protein and a four-turn a-helix (a3), part of which Protein concentration was determined by the method of Bradford (1976)
forms one edge of the nucleotide-binding pocket (Fig. 1). using bovine serum albumin as a reference standard. The molecular weight
Ser-188 is a surface-exposed residue located at the junction of kinesin was taken as 39,700.
of the switch 1 region and the a3-helix, adjacent to the Tubulin was prepared from bovine brain tissue following the protocols
of Ma and Taylor (1995) and stored in a buffer containing 100 mM PIPES,
nucleotide site in kinesin (Fig. 1). The residue is not
1 mM MgCl2 1 mM EGTA, 1 mM Na2GTP,
808C. For microtubule poly-
conserved in kinesin-family motors. To further probe con- merization, thawed tubulin was spun at 100,000 g for 20 min, 48C, to remove
formational changes in the switch 1 region, we have made use aggregated material, and 1 mM Na2GTP 10% DMSO and 2 mM MgCl2 were
of the conservative mutation S188C for SDSL in a previously then added to the supernatant. Polymerization was induced by 30 min
studied kinesin construct containing no other reactive cys- incubation at 378C. Taxol (20 mM) was added and the microtubules were spun
down at 100,000 g for 25 min, 258C. The pellet was resuspended in a buffer
teines (Rice et al., 1999). Previous use of this construct has containing 25 mM NaCl, 25 mM PIPES, 1 mM EGTA, 2 mM MgCl2, 20 mM
been limited to labeling with undecagold to locate the posi- taxol, pH 7.0, 258C. The final concentration of the tubulin ab-dimer was
tion of the catalytic domain of kinesin on the microtubule, 200250 mM. The molecular weight of the dimer was taken to be 110,000.
and a single fluorescence resonance energy transfer distance The microtubule-activated ATPase rates were determined using 1 mM
kinesin, 10 mM microtubules in buffer A containing an additional 2 mM
observation from kinesin already complexed with micro-
Na2ATP. Liberated phosphate was determined using the malachite green
tubules (Rice et al., 1999). There have been no previous spec- assay (Kodama et al., 1986) with ATPase time points taken every 30 s over
troscopic studies directly monitoring conformational changes a 150 s time period. The slope of a least-squares fit to the data gave the
in the switch 1 region upon interaction with MTs. Here we ATPase rate.
demonstrate that EPR probes at the Cys-188 site show a con- First derivative EPR spectra were accumulated with an ER/200D
spectrometer (IBM Instruments, Danbury, CT) with the following in-
formational change in the switch 1 region upon binding to strument settings: microwave power, 25 mW; center field, 345.5 mT;
MTs, and that the conformational change differs in the di- modulation, 0.1 mT at a frequency of 100 kHz; gain, 0.12.0 3 106.
and triphosphate states. Between 10 and 25, 12.7 mT wide, 50 s sweeps were averaged for final data

Biophysical Journal 84(5) 31903196

3192 Naber et al.

analysis. Kinesin (50 mM) in buffer A with added 5 mM Li4AMPPNP, 5 mM

Na2ADP, or 5 mM Na2ADP and either 2 mM AlCl3 or 2 mM BeCl2 and 10
mM NaF was mounted in a 50 mL glass capillary and placed in a TE011
cavity. Temperature was maintained by passing cooled or heated air through
the radiation slits in the cavity and monitored by a thermistor located close to
the experimental sample. For spectra in the presence of microtubules, 30 mM
kinesin and 90 mM MTs were added to ligand-modified buffer A. The mix-
ture was spun at 100,000 g for 25 min, 258C. The pellet was then mounted
in a depression in a rexolite flat cell, surrounded by grease, covered with a
glass coverslip to prevent dehydration, and placed in the TE011 cavity for
data accumulation.
Spectra for MSL-labeled protein in the presence of ADP were decon-
volved by first obtaining the pure components. An almost pure immobilized
spectrum was observed experimentally in the presence of MTs, 28C, and an
almost pure mobile component was observed experimentally in the absence
of MTs, 308C. The small amounts of the other component in each of these
were deleted by subtraction, and the resulting spectra taken as an approx- FIGURE 2 EPR spectra of MSL-labeled KS188C taken at 208C in the
imately pure component for further analysis. presence and absence of microtubules. The horizontal axis is magnetic field
Taxol was obtained from Molecular Probes (Eugene, OR). MSL was (width 12.7 mT), the vertical axis is the derivative of absorption. The top
obtained from Aldrich (Milwaukee, WI). MTSL was obtained from Toronto three spectra are in the presence of 5 mM ATP analog. The bottom spectra
Research Chemicals (Toronto, Ontario, Canada). All other compounds are in the presence of ADP. Locations AC in the top three spectra, and
(nucleotides, phosphate analogs, and buffer salts) were obtained from Sigma locations D and E in the ADP spectrum, are discussed in the text. All spectra
Chemical (St. Louis, MO). (Figs. 25) are normalized with respect to the central peak height to enhance
the visual display of spectral differences in the low-field region.

observed with ADP at the nucleotide site. When MSL-

RESULTS
KS188C ADP binds to MTs, there is a small outward shift of
d

To fix notation, KS188C will refer to the K349 Cys-lite spectral intensity at both the low field (D) and high field (E)
kinesin monomer, with mutation S188C used in these stud- regions of the spectra. This implies that a fraction of the
ies. KS188C has a MT-activated ATPase rate (1 mM probes in the MSL-KS188C ADP state now become more
d

KS188C, 10 mM MT) of 15.4 6 0.4 kinesin
1 s
1 (mean immobilized upon binding to MTs. Fig. 3 shows the results
6 range, two observations). This is ;75% of the value pre- from MSL-labeled protein when the temperature is lowered
viously observed for Cys-lite K349 (Rice et al., 1999) when to 58C. The spectral changes are more pronounced than at
differences in ionic strength are taken into account. With 208C. For the ATP-analog states, there is a significant in-
MSL-labeled protein, MSL-KS188C, the MT-activated ward shift in spectral intensity in the low-field components
ATPase decreases to 9.8 6 0.4 kinesin
1 s
1 (two obser- (A ! B) and a decrease in intensity of the high-field dip (C).
vations), and for MTSL-KS188C the MT-activated ATPase This again implies an increase in mobility of a fraction of the
is 8.3 6 0.4 kinesin
1 s
1 (two observations). These values probes upon binding to MTs. There is, however, a small
imply that spin probe-labeled protein retains catalytic com- outward shift of the high-field dip (C) in the presence of
petence, although the modification of the protein at Cys-188 ADP BeFx, indicating that the magnitude of the mobil-
d

does affect the rates of some step(s) in the cycle, empha- ity increase may be slightly less than with the other two
sizing the importance of this region.
Fig. 2 demonstrates our fundamental results. The spectra
in the absence (presence) of MTs are shown by dashed
(solid) lines for MSL-labeled kinesin at 208C. In the absence
of MTs, the spectra appear independent of the state of
the bound nucleotide. The spectra show that when MSL-
KS188C triphosphate analog binds to MTs, there is a shift of
d

EPR probes from a more immobilized to a more mobile

component. This is demonstrated by the shift in spectral
intensity from the low-field component, A, to an up-field
component, B, and by the decrease in intensity at the high-
field dip, component C, in the AMPPNP, ADP AlFx, and d

ADP BeFx spectra when the spectra in the presence of MTs

d

are compared with those in the absence of MTs. Addition-

ally, the fact that these spectral components appear to occur
at the same positions in the magnetic field suggests that the FIGURE 3 EPR spectra of MSL-KS188C taken at 58C. The top three
probes in all three triphosphate-analog states are reporting spectra are in the presence of ATP analogs. The bottom spectrum is in the
the same conformational change. A small spectral change is presence of ADP.

Biophysical Journal 84(5) 31903196

Kinesin Switch 1 Conformational Change 3193

triphosphate analogs. In the presence of ADP, the spectral

change is smaller with a very small decrease in the mobility
of some of the probes. Upon binding to MTs, there is a slight
outward shift in spectral intensity in the low-field component
(E ! D) as shown by the change in the ratio of peak heights,
and a corresponding outward shift in the high-field dip (F).
These imply a shift to a more immobilized probe orientation.
Figs. 4 and 5 show that the spectral changes obtained in
the presence and absence of MTs when Cys-188 is labeled
with MTSL are more complex. In the absence of MTs at
208C, there is a broad low-field shoulder, component A,
which decreases in intensity in the presence of MTs for the
triphosphate analogs. There is little change in the presence of
ADP (Fig. 4). For all of the nucleotides, there is also an
FIGURE 5 EPR spectra of MTSL-KS188C taken at 58C. The top three
intensity decrease in the right-most portion of the low-field
spectra are in the presence of ATP analogs. The bottom spectrum is in the
shoulder (C), with the appearance of a new peak with greater presence of ADP.
splitting (component B). Using the quantitative analysis and
order parameter of Griffith and Jost (1976), the 4.37 6 0.025
For MTSL-KS188C ADP, spectral differences in the pres-
d
mT (four observations) splitting of this component can be
ence and absence of MTs show only a very small shift to a
modeled as motion within a cone with a vertex angle of 1308.
more immobilized conformation in the presence of MTs.
Cone angles of mobility for the other two, more (A) and less
(C) immobilized components of the spectra, cannot be
defined as precisely. Thus at a minimum, the spectra indicate
The effect of temperature
the presence of three distinct spectral components and
potentially more. As noted, the spectra we have obtained are complex. Many
At 58C (Fig. 5), in the presence of triphosphate analogs, are the sum of three or more distinct spectral components,
MTSL-KS188C shows the same basic spectral change as suggesting transitions between three or more distinct protein
observed at 208C. For the low-field peak, there is a shift in states. These are difficult to deconvolute, confounding more
spectral intensity from a more immobilized component (A) to quantitative analysis. However, for MSL-KS188C bound to
a component (B) of intermediate immobilization. This is MTs in the presence of ADP, the probe shows only two
clearest for MTSL-KS188C, where the A ! E and B ! D different spectral components. Investigating the relative frac-
splittings of 6.64 6 0.05 mT (two observations) and 4.52 6 tions of these two components in the spectra of MT MSL- d

0.04 mT (four observations), respectively, corresponds to an KS188C ADP as a function of temperature allows us to define
d

increase in the cone angle of motion from 528 to 1268, with better the thermodynamics associated with the transition
the larger splitting corresponding to the smaller cone angle. between the two protein conformations. Fig. 6, top, shows the
spectra taken as the immobile component (solid line), and the
mobile component (dashed line), used in defining the spectral
shift as temperature varied for MT MSL-KS188C ADP. Pure
d d

components were obtained as described in Methods. The

spectra of MT MSL-KS188C ADP could then be expressed
d d

as a linear sum of fractions of the two spectra shown in Fig. 6,

top. Fig. 6, middle (MT MSL-KS188C ADP, 158C), shows
d d

such a linear summation. Fig. 6, bottom, shows the quality of

the fit in Fig. 6, middle, giving the residual spectrum obtained
by subtracting the fitted spectrum from the experimen-
tally determined spectrum. Letting K [immobile fraction]/
[mobile fraction] be the equilibrium constant between the two
conformations in the spectra of MT MSL-KS188C ADP as
d d

a function of temperature, Fig. 7 gives
Rln(K) for the

titration data versus 1/T. Using the vant Hoff relationship, the
slope of a least-squares linear fit to the data over the
temperature range 5308C (r 0.99) yields a mean value
FIGURE 4 EPR spectra of MTSL-KS188C taken at 208C. The top three for DH0
96 kJ/mol and a mean DG of 2.2 kJ/mol. This
spectra are in the presence of ATP analogs. The bottom spectrum is in the value for DH0 implies a significant conformational change
presence of ADP. involving more than just a few amino acids. By comparison,

Biophysical Journal 84(5) 31903196

3194 Naber et al.

DISCUSSION
We have examined the changes in mobility of two different
EPR probes covalently bound to KS188C in triphosphate-
analog and diphosphate states in the presence and absence of
MTs. The resulting EPR spectra are complex in that some
have three or more distinct components. To simplify the dis-
cussion of the data, we have assumed that a given spectral
component represents a single probe conformation. How-
ever, we cannot rule out the possibility of multiple confor-
mations giving similar spectral signatures. Thus we may be
underestimating the number of distinct probe conformations
involved. Nonetheless, taken in toto, the data unambiguously
report a conformational change at Cys-188 when kinesin
binds to MTs. Furthermore, the probes report different
changes in mobility between ADP- and ATP-analog states.
AMPPNP and ADP BeFx are ATP analogs with the g-phos-
d

phorous oxygens or fluorines in a tetrahedral orientation.

FIGURE 6 (Top) EPR spectra of MTdMSL-KS188CdADP at 28C (solid ADP AlFx is instead thought to mimic the g-phosphate
d

line) and MSL-KS188CdADP at 308C (dashed line), taken as the immobile hydrolysis transition state intermediate. Both MTSL and
and mobile components of the spectra of MTdMSL-KS188CdADP as MSL show a trend from more immobilized to less im-
a function of temperature. (Middle) Spectrum of MTdMSL-KS188CdADP at mobilized spectra upon binding to MTs for triphosphate
158C (dashed line) and the fit (solid line) as a linear combination of the
analogs. Furthermore, for a given EPR probe at a given
spectra in top panel. (Bottom) Difference spectrum for spectra in middle
panel. temperature, the triphosphate analogs report similar spectra
irrespective of the type of coordination around the g-phos-
phorous. For ADP the situation is reversed. Although small,
the enthalpy of a-helix formation in water is ;5 kJ/mol per the trend is from more mobile to less mobile probes.
residue (Hermans, 1966). It must be emphasized, however,
that we do not know the precise secondary structure involved
in the kinesin conformational change, and that the adjacent RELATIONSHIP TO OTHER STUDIES
protein surface can also significantly influence the enthalpy. A number of crystal structures of kinesin-family motors
For the remaining spectra, a change of comparable magnitude have been solved and are compatible with the possibility of
appears to be occurring, but this cannot be quantitated due to a conformational change at the Cys-188 position. Kinesin
the overlap of the spectral components. Ser-188 is at the junction of the a3-helix and loop 9. Loop 9
and the adjacent a3a-helix at the nucleotide site form the
switch 1 region of kinesin-family motors (Fig. 1). Fig. 8
shows the a3-helix-loop 9 junction (Ca backbone only) from
representative kinesin-family x-ray structures. The a3-helix
is at the top; loop 9 is at the bottom. The residue homologous
to kinesin Ser-188 is shown as a solid sphere. As is evident,
the x-ray structures show considerable lateral motion at the
Cys-188 position. In Kar3 ADP (Yun et al., 2001)/ncd ADP
d d

(Sablin et al., 1996), it is displaced closest to/furthest from

the nucleotide. Other x-ray structures lie between these ex-
tremes. Additionally, in the x-ray structure of Kar3 with the
mutation, R598A (red ), the a3-helix becomes disordered
one amino acid (Asn-582) to the N-terminus (a3-helix side)
of the location homologous to kinesin Ser-188 (Lys-583).
This disorder continues into loop 9.
The switch 1 region of kinesin has a structure homologous
to similar regions in the G-proteins and in the myosin mo-
FIGURE 7 Plot of [immobile fraction]/[mobile fraction] (spectra of Fig.
tors. This structural element was first identified in the G-pro-
6, top) versus 1/T for MSL-labeled MTdKS188CdADP as a function of
temperature. The spectral components representing the mobile and immobile teins as having a conformation that is sensitive to the state of
probes were obtained as described in the Methods. The slope of the least- the nucleotide (Sprang, 1997). It is also sensitive to the
squares linear fit gives DH0. binding of other proteins (Goldberg, 1998; Boriack-Sjodin

Biophysical Journal 84(5) 31903196

Kinesin Switch 1 Conformational Change 3195

Kinesin-family motors generate biologically useful move-

ment only in the presence of the polymer roadway. A
fundamental difference between these previous crystallo-
graphic and molecular dynamics studies, and our present
observations, is that MTs were not present in these earlier
studies. Here we have use EPR spectroscopy to additionally
investigate the influence of the polymer, and show that it
is responsible for a conformational change adjacent to the
nucleotide-binding site. Thus our results support previous
observations of MT-induced conformational changes at the
nucleotide site (Naber et al., 2002). By influencing the struc-
ture of the switch 1 region, the MT could be acting as a nuc-
leotide exchange factor or could be promoting nucleotide
hydrolysis. The rates of both exchange and hydrolysis are
known to be enhanced by the binding of MTs.
In summary, EPR spectroscopy shows a conformational
change at the nucleotide site of kinesin upon binding to MTs.
Thermodynamic analysis implies a significant number of
amino acids are involved.

REFERENCES
FIGURE 8 An overlay of structures (Ca backbone) of representative
kinesin-family motors is shown. The a3-helix is at the top; the contiguous Baker, J. E., I. Brust-Mascher, S. Ramachandran, L. E. LaConte, and D. D.
loop 9 is at the bottom. Loop 9 is part of the switch 1 region. Color coding: Thomas. 1998. A large and distinct rotation of the myosin light chain
kinesin aa 181197, gray (Kull et al., 1996); ncd aa 530546, blue (Sablin domain occurs upon muscle contraction. Proc. Natl. Acad. Sci. USA.
95:29442949.
et al., 1996); kif1a aa 194210, cyan; Kar3 aa 576592, magenta (Yun et al.,
2001); and Kar3-R598A (Yun et al., 2001), red. Kinesin Ser-188 and Baumann, B. A., B. D. Hambly, K. Hideg, and P. G. Fajer. 2001. The
homologous residues (ncd Met-537, kif1a K201, Kar3 Lys-583) are shown regulatory domain of the myosin head behaves as a rigid lever.
Biochemistry. 40:78687873.
as color-coded Ca space-filling spheres. For Kar3dADP with the mutation
R598A (red), the crystal structure becomes disordered one amino acid to the Berger, C. L., and D. D. Thomas. 1994. Rotational dynamics of actin-
a3-helix side of the Lys-583 position. Loop 9 is disordered and the x-ray bound intermediates of the myosin adenosine triphosphatase cycle in
myofibrils. Biophys. J. 67:250261.
structure terminates.
Boriack-Sjodin, P. A., S. M. Margarit, D. Bar-Sagi, and J. Kuriyan. 1998.
The structural basis of the activation of Ras by Sos. Nature. 394:337
343.
et al., 1998). For example, nucleotide exchange factors can Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye
act by inducing a more open switch 1, thereby promoting binding. Anal. Biochem. 72:248254.
nucleotide exchange. In myosin, the switch 1 region covers Cooke, R. 1997. Actomyosin interaction in striated muscle. Physiol. Rev.
the triphosphate moiety of the nucleotide, forming a phos- 77:671697.
phate tube encasing the triphosphates. We have shown that Cooke, R., M. S. Crowder, and D. D. Thomas. 1982. Orientation of spin
nucleotide analogs that are too bulky to fit through the labels attached to cross-bridges in contracting muscle fibres. Nature.
300:776778.
phosphate tube still bind to the nucleotide site of myosin
Fajer, P. G. 1994. Determination of spin-label orientation within the myosin
(Pate et al., 1997), implying that the switch 1 region can head. Proc. Natl. Acad. Sci. USA. 91:937941.
adopt a more open structure. Crystal structure-constrained,
Goldberg, J. 1998. Structural basis for activation of ARF GTPase
electron microscopy reconstructions of S1-decorated actin mechanisms of guanine nucleotide exchange and GTP-myristoyl
filaments have likewise implied an open switch 1 region in switching. Cell. 95:237248.
myosin (Holmes et al., 2002). Gollub, J., C. R. Cremo, and R. Cooke. 1999. Phosphorylation regulates the
The data discussed above led us to explore the possibility ADP-induced rotation of the light chain domain of smooth muscle
myosin. Biochemistry. 38:1010710118.
that the switch 1 region of kinesin-family motors, which in
Griffith, O. H., and P. C. Jost. 1976. Lipid spin labels in biological mem-
all crystal structures is more open than that of myosin, could branes. In Spin Labeling Theory and Applications. L. J. Berliner, editor.
close to cover the phosphates as seen in myosin. We ex- Academic Press, New York. 454523.
plored this hypothesis using molecular dynamics. The Gulick, A. M., H. Song, S. A. Endow, and I. Rayment. 1998. X-ray crystal
simulations indicated the stability of a switch 1 region con- structure of the yeast Kar3 motor domain complexed with Mg.ADP to
2.3 A resolution. Biochemistry. 37:17691776.
formation similar to that found in myosin motors and the
Hermans, J. 1966. Experimental free energy and enthalpy of formation of
G-proteins (Minehardt et al., 2001), with the switch 1 region the alpha-helix. J. Phys. Chem. 70:510515.
directly adjacent to the nucleotide, forming a phosphate tube Holmes, K. C., and M. A. Geeves. 2000. The structural basis of muscle
in kinesin-family motors. contraction. Philos. Trans. R. Soc. Lond. B Biol. Sci. 355:419431.

Biophysical Journal 84(5) 31903196

3196 Naber et al.

Holmes, K. C., J. Kull, and R. Schroeder. 2002. Switch 1 opens on strong Rice, S., A. W. Lin, D. Safer, C. L. Hart, N. Naber, B. O. Carragher, S. M.
binding to actin. Biophys. J. 82:16a. (Abstr.) Cain, E. Pechatnikova, E. M. Wilson-Kubalek, M. Whittaker, E. Pate, R.
Cooke, E. W. Taylor, R. A. Milligan, and R. D. Vale. 1999. A structural
Hubbell, W. L., H. S. McHaourab, C. Altenbach, and M. A. Lietzow. 1996.
change in the kinesin motor protein that drives motility. Nature. 402:
Watching proteins move using site-directed spin labeling. Structure.
778784.
4:779783.
Sablin, E. P., F. J. Kull, R. Cooke, R. D. Vale, and R. J. Fletterick. 1996.
Isas, J. M., R. Langen, H. T. Haigler, and W. L. Hubbell. 2002. Structure Crystal structure of the motor domain of the kinesin-related motor ncd.
and dynamics of a helical hairpin and loop region in annexin 12: a site- Nature. 380:555559.
directed spin labeling study. Biochemistry. 41:14641473.
Sack, S., J. Muller, A. Marx, M. Thormahlen, E. M. Mandelkow, S. Brady,
Kodama, T., K. Fukui, and K. Kometani. 1986. The initial phosphate burst and E. Mandelkow. 1997. X-ray structure of motor and neck domains
in ATP hydrolysis by myosin and subfragment-1 as studied by a modified from rat brain kinesin. Biochemistry. 36:1615516165.
malachite green method for determination of inorganic phosphate.
J. Biochem. (Tokyo). 99:14651472. Schief, W. R., and J. Howard. 2001. Conformational changes during
kinesin motility. Curr. Opin. Cell Biol. 13:1928.
Kull, F. J., and S. A. Endow. 2002. Kinesin: switch I & II and the motor
Serag, A. A., C. Altenbach, M. Gingery, W. L. Hubbell, and T. O. Yeates.
mechanism. J. Cell Sci. 115:1523.
2002. Arrangement of subunits and ordering of beta-strands in an
Kull, F. J., E. P. Sablin, R. Lau, R. J. Fletterick, and R. D. Vale. 1996. amyloid sheet. Nat. Struct. Biol. 9:734739.
Crystal structure of the kinesin motor domain reveals a structural Smith, C. A., and I. Rayment. 1996. X-ray structure of the magnesiu-
similarity to myosin. Nature. 380:550555. m(II).ADP.vanadate complex of the Dictyostelium discoideum myosin
Kull, F. J., R. D. Vale, and R. J. Fletterick. 1998. The case for a common motor domain to 1.9 A resolution. Biochemistry. 35:54045417.
ancestor: kinesin and myosin motors and G proteins. J. Muscle Res. Cell Sprang, S. R. 1997. G proteins, effectors and GAPs: structure and
Motil. 19:877886. mechanism. Curr. Opin. Struct. Biol. 7:849856.
Ma, Y. Z., and E. W. Taylor. 1995. Mechanism of microtubule kinesin Thomas, D. D., R. Cooke, and V. A. Barnett. 1983. Orientation and
ATPase. Biochemistry. 34:1324213251. rotational mobility of spin-labelled myosin heads in insect flight muscle
Minehardt, T. J., R. Cooke, E. Pate, and P. A. Kollman. 2001. Molecular in rigor. J. Muscle Res. Cell Motil. 4:367378.
dynamics study of the energetic, mechanistic, and structural implications Vale, R. D. 1996. Switches, latches, and amplifiers: common themes of G
of a closed phosphate tube in ncd. Biophys. J. 80:11511168. proteins and molecular motors. J. Cell Biol. 135:291302.
Naber, N., R. Cooke, and E. Pate. 1997. Binding of ncd to microtubules Vale, R. D., and R. A. Milligan. 2000. The way things move: looking under
induces a conformational change near the junction of the motor domain the hood of molecular motor proteins. Science. 288:8895.
with the neck. Biochemistry. 36:96819689. Woehlke, G. 2001. A look into kinesins powerhouse. FEBS Lett. 508:291
Naber, N., T. Minehardt, J. Grammer, P. Kollman, R. Car, R. Yount, 294.
R. Cooke, and E. Pate. 2002. The nucleotide site of kinesin-family Yount, R. G., D. Lawson, and I. Rayment. 1995. Is myosin a back door
motors closes upon binding to microtubules. Biophys. J. 82:63a. (Abstr.) enzyme? Biophys. J. 4(Suppl.):4449.
Pate, E., N. Naber, M. Matuska, K. Franks-Skiba, and R. Cooke. 1997. Yun, M., X. Zhang, C. G. Park, H. W. Park, and S. A. Endow. 2001. A
Opening of the myosin nucleotide triphosphate binding domain during structural pathway for activation of the kinesin motor ATPase. EMBO J.
the ATPase cycle. Biochemistry. 36:1215512166. 20:26112618.

Biophysical Journal 84(5) 31903196