19
Conformational changes during kinesin motility
William R Schief* and Jonathon Howard†
Nucleotide-dependent movements of the head and neck of
kinesin have been visualized by cryoelectron microscopy and
have been inferred from single-molecule studies. Key predictions
of the hand-over-hand model for dimeric kinesin have been
confirmed, and a novel processivity mechanism for the
one-headed, kinesin-related motor KIF1A has been discovered.
Addresses
Department of Physiology & Biophysics, University of Washington,
Box 357290, Seattle, Washington 98195-2790, USA
*e-mail: [email protected]
† e-mail: [email protected]
Current Opinion in Cell Biology 2001, 13:19–28
0955-0674/01/$ — see front matter
© 2001 Elsevier Science Ltd. All rights reserved.
Abbreviations
cryo-EM cryoelectron microscopy
FRET
fluorescence resonance energy transfer
Introduction
Some of the most intriguing questions about molecular
motors focus on the conformational changes of these proteins
as they move and do work. These are largely structural questions, such as: what is the sequence of structural states that
produces a net displacement? And, how are the transitions
between those states linked with the chemical steps of ATP
hydrolysis? But more subtle and general issues arise when
the load is considered: how do protein conformational
changes do mechanical work? Do proteins employ a stereotyped ‘power-stroke’ to generate force? What is the role of
thermal fluctuations in force generation?
Atomic structures of the motor proteins and their filaments in all the nucleotide states of the motor’s ATP
hydrolysis cycle would help to understand their action.
For myosin II, atomic structures in different nucleotide
states have provided compelling evidence that the protein
generates movement through a ‘swinging lever arm’
mechanism. For kinesin, the atomic structure has been
solved in only one nucleotide state (ADP bound).
Nevertheless, clues to the conformational changes that
take place during kinesin motility have been obtained
from recent cryoelectron microscopy (cryo-EM), biochemical, and single-molecule experiments. This review
focuses on kinesin results reported during the past year
and touches on parallel results from myosin.
Anatomy of motor proteins
Kinesin and myosin motor proteins appear to operate using
the same basic design. A conformational change in the motor
domain, which is associated with a change in nucleotide
state, is transmitted via a converter domain to a lever domain.
In this way, small conformational changes (on the order of a
few Angstroms) in the nucleotide-binding pocket (which are
associated with binding or hydrolysis of ATP or release of
products) are amplified into the large displacements (which
can be several nanometers) associated with motility.
Kinesin and myosin use a conserved core motor domain,
which is structurally similar to a core region of G proteins,
to produce the initial nucleotide-dependent conformational
change [1]. However, the converter and lever domains of
kinesin and myosin are quite different. Myosin’s lever is an
8 nm light-chain-binding domain that undergoes a large
rigid-body rotation, as seen when myosin is crystallized in
different nucleotide states [2,3,4•]. Evidence that the
myosin lever arm rotates during motility is reviewed by
Geeves and Holmes [5•]. Recent fluorescence resonance
energy transfer (FRET) measurements report a large,
nucleotide-dependent angle change in myosin’s lever arm
[6•] and thereby provide additional support for the swinging
lever arm model for myosin.
Kinesin (Figure 1) has no single domain corresponding to
myosin’s lever, and the inability of one-headed kinesin to
step suggests that movement of dimeric kinesin relies partly
on mechanical amplification by the second (tethered) head
[7,8•]. However, for the free head to reach the next binding
site (8 nm away) while the other head remains bound, a large
conformational change is required to alter the relative positions of the two heads: either the coiled-coil dimerization
domain (red and blue helices in Figure 1) opens up [9,10] or
the ‘neck linker’ (a stretch of ~15 amino acids that links the
conserved motor domain to the coiled-coil stalk [red and
blue spheres in Figure 1]) peels away from the surface of the
motor domain [11•], or both [12]. Thus a number of
domains — the motor, the neck linker, and the dimerization
domain — probably contribute to kinesin’s lever.
Directionality and the converter domain
What determines the direction of a motor protein along its
filament? The concept of a converter domain that acts as a
sort of gear box is useful for understanding the directionality of both myosin- and kinesin-related motors. For
myosin, the ‘converter’ domain forms the junction
between the motor domain and the light-chain-binding
domain (the lever): a change in the nucleotide bound in
the active site alters the structure of the converter domain
[13], and this, in turn, causes rotation of the lever [4•]. The
connection between the active site and the converter
domain is made by the relay helix (also called the switch II
helix), which may couple the opening of the nucleotidebinding pocket to rotation of the converter and the
attached lever [4•,14•]. Myosin VI has a large insert in the
converter domain, and this motor protein moves towards
the minus end of actin filaments, in the opposite direction
from all other characterized myosin motors [15••].
20
Cytoskeleton
domain of myosin [11•,16]. Support for this idea first came
from studies of chimeras of kinesin and the kinesin-related
motor protein ncd (reviewed in references [17–19]). When
the conserved motor domain of ncd (i.e. the region of ncd
that is highly conserved among all kinesin-related proteins)
is put into the kinesin ‘body’, the resulting chimera is plusend directed, like kinesin. Conversely, when the conserved
motor domain of kinesin is put into the ncd body, the
chimera is minus-end directed, like ncd. Truncation experiments with kinesin [20] place the direction-determining
region in the neck linker domain (Figure 1). Although ncd has
no neck linker, the direction-determining sequences of ncd
are also located outside of the conserved motor domain, in the
coiled-coil domain just amino terminal to the motor domain.
Mutations in this domain, called the ncd neck, can reverse the
direction of movement. Remarkably, change of a single
residue in the ncd neck creates a motor that can alternate
between plus-end- and minus-end-directed states [21•].
Figure 1
(a)
In summary, considerable evidence supports the idea that
a ‘converter’ domain can modulate directionality in both
kinesin- and myosin-related motors. More research is
needed, however, to reveal the structural mechanisms that
underlie this directional control and to answer questions
such as how does a molecular gear box really work.
(b)
Processivity and the hand-over-hand model
+
8 nm
Current Opinion in Cell Biology
Crystal structures of dimeric kinesin containing ADP [9], oriented on a
microtubule in a manner similar to the docking proposed by Hirose
et al. [33•]. (a) Shows the rear view of kinesin with the plus end of the
microtubule going into the page and includes a schematic cross
section of two protofilaments and includes a schematic cross section
of two protofilaments. (b) Shows the side-view with the plus end to the
right. The core motor domains (conserved among all kinesin-related
proteins) are gray (residues 8–324). The amino termini are black
(residues 1–7); ADP is yellow; the neck-linkers (residues 325–338)
are displayed as spheres in red (attached head) and blue (tethered
head), and the α-helices of the coiled-coil dimerization domain
(residues 339–372) follow the same red and blue scheme. Only the
first 5 nm of the dimerization domain is visible in this structure. The
‘neck’ region includes both the neck linker and the first 20–30
residues (~3.5 nm) of the coiled-coil. An open question is whether or
not the amino terminus impacts the conformation of the neck. Kinesin
was rendered using the program RasMol.
For kinesin-related proteins, the ‘neck’ region that forms the
junction between the head and dimerization domain
(Figure 1) probably plays the same role as the converter
The processivity of molecular motors, which is the ability
to make multiple steps along a filament before dissociation, varies widely. Kinesin is a highly processive motor
that can take hundreds of steps before falling off the microtubule (reviewed in [16]), whereas the kinesin-related
protein ncd is non-processive and cannot manage even two
steps in succession [21•,22••,23••]. Myosin V moves processively along actin filaments [24•,25•,26••,27•,28•],
whereas muscle myosin II is non-processive [29]. These
differences are thought to reflect different biological functions: motors that operate in large ensembles to drive rapid
movements (e.g. muscle myosin) should spend only a
small fraction of their mechanical cycle attached to the filament and thus be non-processive, whereas motors that
operate alone or in small numbers to transport organelles
(e.g. kinesin) should be processive [29].
What are the properties of a motor protein that allow it to
be processive? In the case of kinesin, recent work from
several groups supports a model [30] in which the two
heads move in a coordinated, hand-over-hand fashion. A
crucial aspect of this coordination was recently demonstrated, namely that the detachment of one head is
contingent on the attachment of the other [7,8•,31•]. For
conventional kinesin and Neurospora kinesin, binding of
the detached, tethered head to the microtubule accelerates
the ATPase and detachment rates of the bound head by a
factor of 10 or more [8•,31•]. The physical basis of this
coordination between the heads is thought to be the
intramolecular strain generated upon binding of the forward
head, although direct evidence for this is still lacking.
Conformational changes during kinesin motility Schief and Howard
21
Figure 2
The proposed mechanochemical cycle of
kinesin, with possible branches indicated.
An animation of this cycle can be found at
www.current-opinion.com/subjects/
jcel/mov1.mov. The neck linker and
dimerization domain are red and blue, as in
Figure 1. Black denotes one side of the
motor domain (the side view of the attached
head in Figure 1b), and gray denotes the
other side. Motor domains are labelled with
their nucleotide state: D represents ADP, T
represents ATP, DP represents ADP·Pi; and
φ represents no nucleotide. Nucleotides are
shown as colored spheres: ADP, yellow;
ATP, green; ADP·Pi, blue. The microtubule
protofilament is schematized as light gray
(α subunits) and dark gray (β subunits)
hemispheres. State 0 represents the crystal
structure with ADP bound [9]; in this
structure the neck-linker clings to the motor
core. The heads are related by a rotation of
120° about an approximately vertical axis, so
if one imagines the black head to lie in the
plane of the page, the gray head is actually
rotated into the page by approximately 60°.
In state 1, one head attaches and rapidly
releases its ADP; the tethered head does
not attach in the absence of added
nucleotide [31•,40•,55]. The dimerization
domain is depicted as if it were attached to
a small load, so it points backward rather
than out of the page as in Figure 1b. In
state 2, the attached head binds to ATP,
causing a conformational change that moves
the tethered head from pointing into the
page to pointing out of the page [33•,43,44]
and advances the load by 1–2 nm
(see ‘load-dependent steps’). Here the
conformational change is depicted as a
movement of the neck-linker on the surface
of the attached motor (red linker moves from
one position to another), which causes an
opening in the lower ~2.5 nm (two heptads)
of the coiled-coil dimerization domain [56]
and thereby allows the tethered head to
rotate and fluctuate about its neck. In
state 3, ATP hydrolysis in the attached head
usually precedes and accelerates
attachment of the tethered head (see text),
but no information is available on the
specific conformational changes due to
hydrolysis (here represented by the average
position of the tethered head moving closer
to the forward microtubule-binding site).
Also, two-headed binding is possible
without hydrolysis [31•,55] (left branch,
states 3→4). In state 4, a conformational
fluctuation of the neck linker has allowed the
tethered head to attach to the microtubule
and become the forward head. In state 5,
the forward head releases its ADP rapidly
following attachment, thus binding more
tightly to the microtubule [8•,31•,40•,50],
trapping its neck-linker in a peeled-away
configuration, and causing intramolecular
strain. In state 6, sustained strain induces
0
ADP
ATP
ADP·Pi
D
D
D
1
φ
D
Both heads bound
before hydrolysis?
2
T
D
T
D
3
DP
T
φ
4
DP
D
Phosphate release
from attached, rear head?
5
ATP binding before
rear-head release?
φ
DP
φ
D
P
T
DP
DP
T
6
φ
D
T
Futile
hydrolysis?
D
7
φ
D
D
φ
0 nm 8
Current Opinion in Cell Biology
detachment of the rear head, owing to the
weaker attachment in ADP·Pi compared to
no nucleotide [31•,40•,50]. ATP binding in
the forward head might cause additional
strain and hence further accelerate rearhead release [11•] (left branch,
states 6→7). The rear head probably
detaches before releasing its phosphate
(see text), but earlier phosphate release is
possible [31•,40•] (right branch). States
6→7; the newly detached head swings
forward, and the load may move forward,
as the strain of the two-heads-bound state
is relieved. We hypothesize that phosphate
release causes the neck linker of the
detached head to shift back to the
ADP position, allowing the re-coiling of the
dimerization domain. Kinesin has now
advanced by 8 nm with respect to state 1,
and the coiled-coil has gained a half twist
(red α-helix and neck linker are now in front
of blue). Whether the next step will reliably
relieve this half-twist is a question that
remains to be answered.
22
Cytoskeleton
A key prediction of the hand-over-hand model is that the
two heads of kinesin can bind simultaneously to the microtubule. Two-headed binding of kinesin to microtubules has
been demonstrated by cryo-EM [10,32], although other laboratories see predominantly one-headed binding under
similar conditions [33•]. This apparent discrepancy might
be accounted for by differences in motor concentration
(higher motor concentrations favor one-headed binding)
and/or ADP concentration (ADP, even in the micromolar
range, is likely to displace one head from the microtubule,
owing to its very high affinity for a free head). Direct
mechanical evidence has now been obtained for a twoheads-bound configuration [34]. Single dimeric kinesin
molecules attached to beads and an optical trap were used
to measure the unbinding force and mechanical stiffness of
the kinesin–microtubule complex. In the presence of saturating concentrations of the nonhydrolyzable ATP analog
AMP-PNP both the unbinding force and the stiffness were
twice that measured in the absence of nucleotide when the
trap pulled the bead toward the plus end of the microtubule.
This strongly suggests that two-headed binding predominates in the presence of AMP-PNP, whereas one-headed
binding predominates in the absence of nucleotide.
Although the evidence presented on the processivity of
KIF1A is rather compelling, these results are not without
controversy, as the findings have been made in only one
laboratory, and other groups have found evidence that the
monomeric motor unc104, the KIF1A homologue in
Caenorhabditis elegans, is not processive [38]. One possible
source for the discrepancy is the particular motor constructs used: the ‘processive’ KIF1A was actually a
truncated form of KIF1A (351 amino acids) fused to
23 amino acids of the neck region of kinesin, and the ‘nonprocessive’ unc104 was a longer construct with 651 amino
acids from unc104 fused to GFP (green fluorescent protein). It is important to confirm the processivity of the
KIF1A constructs in other laboratories and to test other
constructs. Furthermore, although experiments with
KIF1A knock-out mice implicated this monomeric motor
in the transport of vesicle precursors [39], it remains to be
determined whether KIF1A’s diffusion-based mechanism
can still work against the high loads expected when the
motor is challenged to move organelles through the crowded
cytoplasm (for further discussion of KIF1/UNC104 see the
review by Bloom, pp 36–40 of this issue).
Kinesin’s mechanochemical cycle
Processivity of a one-headed motor
Although evidence indicates that the two heads of kinesin
alternate in a hand-over-hand stepping fashion, a novel processivity mechanism appears to operate in the one-headed,
kinesin-related motor protein KIF1A. Recent work surprisingly indicates that KIF1A is processive: on average, each
encounter of KIF1A with a microtubule results in a net displacement of about 1 µm toward the plus end [35••]. Like
dimeric kinesin, monomeric KIF1A uses two tubulin-binding motifs to maintain its association with the microtubule,
but unlike kinesin, which has only one motif per head,
KIF1A has two different motifs in the one head. One of
these motifs is conserved among all kinesin-related proteins but the second motif is not. The second motif is a
lysine-rich loop (the ‘K loop’) that interacts with the glutamate-rich carboxyl terminus of tubulin [35••,36•,37•], and
deletion or mutation of this loop results in loss or reduction
of processivity [36•,37•]. In addition, a non-processive
motor could be converted into a processive one by extending the K-loop of a conventional kinesin monomer with six
lysines [37•]. Experiments with microtubules whose tubulin subunits were enzymatically digested at the carboxyl
terminus identified the carboxy-terminal region of tubulin
as the binding partner of the K loop [37•].
The movement of KIF1A differs from kinesin in that its
velocity fluctuates wildly. This suggests that diffusion
along the microtubule plays an important role in motility
[35••,37•]. Consistent with a diffusional mechanism, the
coupling between nucleotide hydrolysis and stepping is
loose: several ATP molecules are hydrolyzed per 8 nm
step even in the absence of load. This contrasts with conventional kinesin, which hydrolyzes one ATP per 8 nm
step at low load (see below).
Whether a motor protein is processive or not, the motor’s
filament-binding sites must alternate between at least two
mechanical states (attached to filament and detached), and
the nucleotide-binding pockets must cycle through four
chemical states (empty, ATP, ADP·Pi and ADP). To understand the mechanism by which a motor protein moves, one
must define the dominant pathway(s) that the motor follows through all possible mechanochemical states. All
pathways occur with some non-zero probability, and the
most frequently followed path may change depending on
the external load, nucleotide concentration, temperature,
pH, and salt concentration. Here we discuss the dominant
path of kinesin under conditions typical of a laboratory
motility or biochemical assay: low load (<1 pN), saturating
ATP (~1 mM) and negligible ADP and Pi (~nM), room
temperature, and a pH of 7.
Several models for kinesin’s dominant mechanochemical
cycle have been proposed recently [8•,11•,14•,31•,40•,41].
Although some aspects of the proposed pathways are
agreed upon, there are several important differences
among the models. First we will review those aspects that
are widely accepted, then in a separate section below we
will delve into the controversial issues. To frame the discussion, we propose a mechanochemical cycle for kinesin
in Figure 2 (central panel) that extends the model in reference [8•], and we highlight possible branches from the
main path (side branches in Figure 2).
A key early finding was that microtubules accelerate the
release of ADP, which binds tightly to both heads of kinesin
in solution. But when microtubules are mixed with dimeric
kinesin, ADP dissociates from only one head unless ATP is
present in the solution (Figure 2, states 0→1) [42]. It is
Conformational changes during kinesin motility Schief and Howard
therefore widely believed that binding of ATP to the
attached head precedes a conformational change that allows
the other (tethered head) to attach and release its ADP
(Figure 2, states 2→5), although the details of this conformational change are a matter of debate. The presence of
the non-hydrolyzable ATP analogue AMP-PNP also accelerates release of the second ADP, showing that hydrolysis is
not necessary for ADP release from the second head
(Figure 2, left branch, states 3→4). However, as argued
below, hydrolysis usually takes place before attachment of
the second head (main path of Figure 2, states 3→4).
Cryo-EM of kinesin-decorated microtubules has also provided hints about the movements of kinesin. Comparison
of microtubules decorated with kinesin dimers in the
absence of nucleotide and in the presence of AMP-PNP
shows that the binding of AMP-PNP is associated with a
large conformational change that moves the tethered head
from the left of the attached head to the right of the
attached head, where the plus end of the microtubule is up
([33•,43,44]; see Update, but see also [10] for another interpretation). This conformational change appears to take
place within the neck-linker of the attached head, because
a similar structural change is seen when AMP-PNP binds
to a kinesin monomer whose neck-linker has been labeled
near its carboxyl terminus with a gold particle [11•]. These
electron microscopy studies thus support the notion that
the tethered head acts like a lever and the neck acts like a
converter domain (and possibly also part of the lever).
Several other important aspects of kinesin’s pathway in
Figure 2 have been agreed upon. One is the fuel efficiency:
kinesin is a ‘tightly coupled’ motor, it hydrolyzes one ATP
molecule for each 8 nm step under low load
[45•,46•,47,48,49•]. Optical trapping data suggest that the
coupling remains tight even under loads as large as 5 pN
[49•]. A second feature is the maximum motor speed: conventional kinesin takes ~100 steps per second at saturating
ATP and low load (so moving 8 nm from state 1 to state 7
in Figure 2 takes ~0.01 s). Another widely agreed feature of
kinesin motility is that the attachment between the motor
domain and the microtubule is nucleotide dependent: the
motor dissociates slowly (‘binds strongly’) with no
nucleotide or AMP-PNP bound and dissociates more
quickly (‘binds weakly’) with ADP or ADP·Pi bound
[8•,31•,40•,50]. These differences in binding strength
ensure that when both heads of kinesin are attached to the
microtubule the rear head with ADP·Pi is more likely to be
released than the forward head with no nucleotide bound
(Figure 2, states 5→6). Finally, it has recently become clear
that kinesin possesses an internal conformational switch
that prevents motility unless cargo is bound to its carboxyl
terminus, termed ‘tail domain’. Without cargo (which can
be an organelle, bead, or glass surface, for example) bound
to the carboxyl terminus, kinesin folds up so that the tail
binds to the motor domain and inhibits ATP hydrolysis
[51•,52]. So, the pathway in Figure 2 only applies to kinesin
with cargo bound, or with the tail domain deleted.
23
Uncertainties regarding kinesin’s
mechanochemical cycle
Many important aspects of the conformational changes during kinesin motility are not understood. Given that ATP
binding to the attached head of kinesin induces a positional
change in the tethered head, one important question is:
how is the initial conformational change in the nucleotidebinding pocket of the attached head transmitted and
amplified into a larger motion of the tethered head?
As the neck-linker and dimerization domain provide the
physical connection between the heads, recent work [11•]
has focused on changes within kinesin’s neck-linker upon
ATP binding. Electron paramagnetic resonance (EPR)
spectroscopy of probes attached to the neck-linker suggests that this region is highly mobile in the presence of
ADP and in the absence of nucleotides when the motor is
in solution or attached to a microtubule. However, the
neck-linker is immobile in the presence of AMP-PNP
when the motor is attached to a microtubule [11•]. These
observations have led to a model in which ATP binding
induces a disordered-to-ordered conformational change of
the neck-linker [11•,14•,53]. However, this model appears
to be at odds with a battery of existing data that support a
rather different conclusion. A highly disordered neck in
the ADP state is inconsistent with the ordered conformation in the dimer crystal structure [9] (Figure 1), the lack of
two-fold symmetry in the dimer solution structure ([54];
see also Update) (the coiled-coil is symmetric, so a disordered neck should lead to two-fold symmetry of the
heads), and the reasonably well ordered conformation seen
in the cryo-EM studies of dimeric kinesin mentioned
above ([33•,44] and Update). Indeed, the observation
using cryo-EM of a reduced mass of the tethered head in
the presence of AMP-PNP [33•,43,44] is consistent with
the tethered head having greater conformational freedom
when ATP is bound to the attached head compared to
when no nucleotide is bound. The ADP-release experiments mentioned above [42] are also more consistent with
an ordered neck in the ADP state: if the ADP state were
highly disordered, one would expect that the tethered head
would be able to attach to the microtubule and release its
ADP even in the absence of ATP. The EPR studies of Rice
et al. [11•] were performed on severely truncated
monomeric motors whose ATPase cycles are poorly coupled to motility (many ATPs hydrolyzed per mechanical
step). Thus, it is possible that the high mobility of the
neck-linker in these constructs arises from this truncation.
The results from cryo-EM [11•,33•,43,44] and ADPrelease experiments [31•,42,55] are consistent with a
model in which ATP binding to the attached head results
in greater mobility of the tethered head, in contrast to the
conclusion of Rice et al. The proposal by Schnitzer et al.
[41] of an isomerization following ATP binding, in which
kinesin undergoes rapid fluctuations between two conformational substates in equilibrium, is also consistent with
ATP-induced conformational freedom of the tethered
24
Cytoskeleton
head. Finally, a disordered conformation of the ATP state
may explain the lack of a kinesin–ATP crystal structure.
How could ATP binding to the attached head increase the
mobility of the tethered head? ATP binding to the
attached head might result in a partial opening of the
coiled-coil dimerization domain, as depicted in Figure 2
(state 2). In this scenario, ATP binding causes the necklinker of the attached head to shift from one ordered
conformation to a second that is incompatible with a fully
closed coiled-coil (Figure 2, states 1→2). In fact, the first
segment (~2.5 nm) of kinesin’s coiled-coil departs from
the classical leucine zipper pattern and has the potential to
uncoil [56]. In summary, although there is considerable
evidence that ATP binding causes a forward-directed
motion of the tethered head, the details of this conformational change remain to be elucidated. Single molecule
studies have provided additional insight into changes associated with ATP binding and these are discussed in the
section on load-dependent steps.
Currently, we have no direct information on the conformational changes associated with ATP hydrolysis, but
biochemical data allow us to address the important question
of whether ATP hydrolysis in the attached head takes place
before or after attachment of the tethered head (Figure 2).
Several lines of evidence indicate that hydrolysis precedes
attachment (Figure 2, states 2→4): first, ATP binding to the
attached head accelerates the release of the fluorescent
ADP-analog mantADP from the detached head several-fold
(a factor of 3–12) more effectively than AMP-PNP [31•,55].
Second, the rate constant for ATP hydrolysis by the kinesin
monomer attached to a microtubule (which may be approximately the same as the rate constant for hydrolysis by the
attached head of the dimer) is more than six-fold higher than
the rate constant for release of mantADP from the tethered
head in the presence of AMP-PNP (200 s–1 [57] versus 30 s–1
[55]). Third, the rate constant for ATP hydrolysis is likely to
be even greater in the attached head of the dimer, in order to
reconcile the low affinity of the attached head for AMP-PNP
(Kd ~ 1 mM [50,55]) with the high apparent second order
association rate (or ‘on-rate’) for ATP binding (1–2 µM–1 s–1
[55]). (The low affinity and high apparent on-rate indicate that binding of ATP to the attached head is followed by a very rapid step, with a rate constant of
k = Kon-apparent × Kd ≈1000–2000 s–1, that does not occur after
binding of AMP-PNP and is therefore probably hydrolysis.)
The above data suggest that hydrolysis in the attached
head usually precedes and accelerates the binding of the
tethered head, as shown in Figure 2 (states 2→4), implying
that the hydrolysis step is associated with a productive conformational change. On the contrary, Rice et al. [11•] find
that a kinesin mutant defective in ATP hydrolysis (E236A)
is nevertheless able to accelerate tethered head ADP
release to the same rate as wild-type (Figure 2, left branch,
states 3→4), and they conclude on that basis that ATP
binding alone drives forward stepping. Evidently the
E236A motor with bound ATP has different properties
than the wild-type motor with bound AMP-PNP. Because
other kinetic properties of this mutant are aberrant (ADP
release is 100 times faster than in wild-type [11•]), we do
not think it provides a good model for the ATP state.
The next chemical step following the hydrolysis of ATP,
the release of phosphate from the rear head, is another
important area of debate. Hancock and Howard [8•] argue
that detachment of the rear head is a prerequisite for phosphate release, on the basis of their findings with a
one-headed heterodimer (Figure 2, states 5→7). Rice et al.
[11•] propose, without evidence, that a one-headed
monomer detaches after phosphate release so that attachment and detachment take place in the same nucleotide
state (i.e. with ADP bound). However, this model is probably incorrect because it fails to account for the
single-motor motility observed with monomers in the
assays of Young et al. [58]. In these assays, single monomeric
motors anchored to glass coverslips could propel microtubules, even though the motors could not remain
continuously attached to the microtubules, because
methylcellulose prevented microtubule escape from the
surface. In the model of Rice et al. [11•], the monomers
would attach and detach in the same conformation, so a
single monomer could not move a microtubule. This problem does not exist if detachment takes place before
phosphate release, as argued by Hancock and Howard [8•].
In this case, the bound phosphate prevents premature
release of ADP and subsequent futile hydrolysis cycles in
the attached rear head (Figure 2, right branch), helping to
ensure that more than 90% of the ATP molecules
hydrolyzed result in a mechanical step [45•]. Cross and colleagues [40•] postulate that phosphate might be released
just before rear-head detachment (Figure 2, right branch).
Direct measurements of the rear-head phosphate release
kinetics are necessary to resolve this issue with certainty.
A final area of uncertainty concerns the release of the rear
head from the microtubule. It is clear that detachment of
the rear head is contingent upon attachment of the forward
head (Figure 2, states 3→6) [8•,31•] but it is not known
whether the intramolecular strain generated by the simultaneous binding of both heads is sufficient to pull the rear
head off the microtubule [8•,31•], or whether subsequent
ATP-binding to the forward head is required as well
[11•,14•]. Interestingly, if ATP binding is strictly required
(Figure 2, left branch, states 6→7) then the first 8 nm step
of kinesin on a microtubule will require two ATPs.
Rationale for kinesin’s cycle
While debates on the exact mechanochemical pathway (or
pathways) of kinesin continue, is there a simple way to
rationalize the mechanochemical coordination between
the two heads? An intriguing possibility is that the symmetry of kinesin may provide a structural basis for the
communication between the heads [16]. Most homodimeric proteins have an (approximate) axis of two-fold
Conformational changes during kinesin motility Schief and Howard
symmetry due to the isologous interaction of the
monomers [16]. This means that the two heads are related
by a 180° rotation. Interestingly, the rat kinesin homodimer (with ADP bound to both heads; Figure 1) deviates
more from symmetry than any other homodimeric protein,
to our knowledge. Nevertheless, the heads are still related
by rotation of ~120°, so binding both heads to a translationally symmetrical microtubule will require very
considerable intramolecular strain when the heads are in
the same nucleotide state [9]. It is therefore reasonable to
hypothesize that, whenever both heads of kinesin are in
the same nucleotide state, simultaneous binding to a
microtubule will be difficult (i.e. a fluctuation large enough
to reach such a highly-strained state will be rare).
It follows that the two heads of kinesin might only bind
simultaneously to the microtubule if they are in different
nucleotide states; conformational differences between the
heads, presumably in the neck regions, might then facilitate the achievement of the required translational
symmetry by a directed conformational change or a thermally induced fluctuation (e.g. Figure 2, transition from
state 3→4). We suggest a simple symmetry rule that helps
account for the mechanochemical path of the kinesin dimer
shown in Figure 2: there are two classes of states, those with
phosphate (ATP, ADP·Pi) and those without phosphate
(nucleotide free and ADP), and two-headed binding is only
allowed when the heads are in different classes.
Load-dependent steps
To understand how motor proteins do mechanical work
we need to resolve which of the motor’s chemical steps
are associated with mechanical displacements, and determine the size of each displacement. Optical trapping
experiments have revealed that kinesin makes 8 nm displacements (e.g. [49•]). But is the entire 8 nm
displacement of kinesin accomplished within a single
chemical step such as ATP binding? Or do several chemical steps each contribute incremental advances
(substeps) that add up to 8 nm? The chemical steps associated with displacements will be slowed down by
opposing external loads; these are the load-dependent
steps. From a functional point of view, the important
question is which steps are rate limiting (i.e. limit the
speed) when the motor is challenged by a load?
What is known so far about the load-dependent steps of
kinesin? Mechanical experiments are consistent with
ATP binding being load dependent. The key observation is that at very low ATP concentrations, when ATP
binding is rate limiting, kinesin slows down under even
a small load. If ATP binding is irreversible (or if it is
reversible, but followed immediately by an irreversible
step such as hydrolysis) then this observation implies
that the on-rate for ATP binding is load dependent.
Alternatively, if ATP binding is reversible and is followed by rapidly reversible steps, then the load
dependence could be distributed over any of these steps.
25
For example, in the scheme of Schnitzer et al. [41], the
load dependence is attributed to a rapidly reversible
isomerization of the ATP-bound state.
An estimate for the displacement associated with ATP
binding can be made using Boltzmann’s law as follows. If
the transition state for the ATP-binding step corresponds
to moving the load through a distance d relative to the
nucleotide-free state, then application of an opposing force
of magnitude F will reduce the rate constant for the ATPbinding step by the factor exp(–Fd/kT), where k is the
Boltzmann constant and T the absolute temperature. If
this step is irreversible, then at low ATP concentration
kinesin’s velocity can be modeled as v(F) = v0exp(–Fd/kT),
where v(F) and v0 are the velocities of loaded and
unloaded kinesin, respectively. In the low-load limit, the
slope of the velocity–load curve is –v0d/kT. This expression
for the slope can be compared to measured velocity–load
curves to extract a value for d. From velocity–load curves
of kinesin measured at low levels of ATP using optical
traps [49•,59,60] or glass microneedles [61] the slope at low
load falls in the range –v0/(2.5 pN) to –v0/(7 pN), so we
find that d ≈ 0.6–1.6 nm. Thus ATP binding (or steps that
are in rapid equilibrium with the ATP-binding step) is
associated with a conformational change of some 1–2 nm.
If ATP binding is associated with an advance of only
1–2 nm, is there evidence for additional load-dependent
steps that are rate limiting? First, the shape of the velocity–load curves measured for kinesin are inconsistent
with a single load-dependent step. All the velocity–load
curves measured for kinesin have been approximately
linear or slightly curved downward [49•,59–63]. These
shapes indicate that there are at least two load-dependent steps: one that is rate limiting at low load and
another that is rate limiting at high load [64]. Additional
evidence for a second load-dependent transition in
kinesin’s cycle comes from experiments showing that
increasing the load causes both the motor speed and the
‘effective’ rate of ATP binding to decline [49 •].
However, precisely which chemical transitions are rate
limiting under a given load remains to be determined. A
combination of experiments and modeling [41,64,65]
will be necessary to make these determinations.
‘Substeps’ of kinesin’s 8 nm step might be detectable
experimentally. By lowering the ATP concentration and
raising the load, it might be possible to slow down the two
different steps: an ATP-binding step and a second step
with stronger load dependence. A plausible candidate for
the second step is the release of the rear head from the
microtubule (Figure 2, states 6→7). When the rates of the
two steps are close, they will both be rate limiting, and in
principle the displacements associated with both steps
should be detectable with the typical time resolution of
~1 ms for an optical trap. Alternatively, the absence of substeps at low ATP levels and high load will indicate that
ATP binding has the strongest load dependence.
26
Cytoskeleton
Conclusions
Over the last few years, a general picture of how motor proteins move and generate force has emerged: small conformational changes initiated in the nucleotide-binding pocket of
the motor domain are amplified into large structural changes
in the protein. For kinesin, the details of these conformational changes are becoming clearer. Biochemical, cryo-EM,
and single-molecule experiments all indicate that a major
conformational change takes place upon ATP binding to the
attached head. This change permits the attachment of the
tethered head to the microtubule and is associated with a
movement of the load towards the plus end of the microtubule. Hydrolysis probably accelerates attachment of the
tethered head in the course of kinesin motility. The attachment of the tethered head is a prerequisite for the release of
products: first of ADP from the forward head, and then of
phosphate from the rear head. The coupling between stepping and nucleotide hydrolysis is tight because the chemical
steps are prerequisite for mechanical steps and mechanical
steps are necessary for chemical steps. Further clarification of
the conformational changes associated with kinesin motility
may come from several directions, including a crystal structure of kinesin in the ATP state (comparison with the ADP
structure may reveal kinesin’s ‘swinging lever arm’), direct
measurement of the relative motions of the two heads while
stepping, recording substeps, and measuring the elasticity of
kinesin. Careful modeling that incorporates the available biochemical, biophysical, and structural data will also be important to further elucidate kinesin’s mechanism of motion.
2.
Rayment I, Holden HM, Whittaker M, Yohn CB, Lorenz M, Holmes KC,
Milligan RA: Structure of the actin-myosin complex and its
implications for muscle contraction. Science 1993, 261:58-65.
3.
Dominguez R, Freyzon Y, Trybus KM, Cohen C: Crystal structure of a
vertebrate smooth muscle myosin motor domain and its complex
with the essential light chain: visualization of the pre- power
stroke state. Cell 1998, 94:559-571.
4.
•
Houdusse A, Kalabokis VN, Himmel D, Szent-Gyorgyi AG, Cohen C:
Atomic structure of scallop myosin subfragment S1 complexed
with MgADP: a novel conformation of the myosin head. Cell 1999,
97:459-470.
The authors argue that the new conformation corresponds to a detached,
stable ATP state that occurs between the rigor (post-power stroke) and
transition (pre-power stroke) states.
5. Geeves MA, Holmes KC: Structural mechanism of muscle
•
contraction. Annu Rev Biochem 1999, 68:687-728.
A comprehensive review of the mechanism of myosin motility.
6.
•
Shih WM, Gryczynski Z, Lakowicz JR, Spudich JA: A FRET-based
sensor reveals large ATP hydrolysis-induced conformational
changes and three distinct states of the molecular motor myosin.
Cell 2000, 102:683-694.
FRET experiments report a nucleotide-dependent ~70° rotation of the
myosin lever arm that is complementary to and consistent with previous
crystallographic data. The authors provide evidence for two distinct ‘prestroke’
angles of the lever.
7.
8.
•
Hancock WO, Howard J: Kinesin’s processivity results from
mechanical and chemical coordination between the ATP
hydrolysis cycles of the two motor domains. Proc Natl Acad Sci
USA 1999, 96:13147-13152.
By comparing the biochemical cycles of one-headed and two-headed kinesin,
the authors show that, in dimeric kinesin, release of the rear head is contingent upon attachment of the forward head. Evidence is provided that the rear
head contains ADP·Pi when it is released. A mechanochemical pathway for
kinesin is proposed that is consistent with a wide body of data.
9.
Update
Single-molecule recordings on myosin II constructs with
genetically engineered levers shows that the workingstroke scales with the length of the lever [66], providing
strong support for the swinging lever-arm model. A recent
study employing small-angle X-ray and neutron scattering
provides strong evidence that the structure of dimeric
kinesin in solution is consistent with the asymmetric
conformation in the ADP crystal structure [67•].
Cryo-EM of the microtubules decorated with dimeric
kinesin-related proteins confirms that these proteins, like
kinesin, attach by one head only. In the absence of
nucleotides, the tethered head of Neurospora kinesin can
occupy either one or two positions, leading the authors to
suggest that either head of the asymmetric dimer can bind to
the microtubule [68].
Acknowledgements
We appreciate comments on this manuscript from the members of the
Howard Laboratory. We thank Linda Amos for a suggestion on Figure 1.
References and recommended reading
Papers of particular interest, published within the annual period of review,
have been highlighted as:
• of special interest
•• of outstanding interest
1.
Kull FJ, Sablin EP, Lau R, Fletterick RJ, Vale RD: Crystal structure of
the kinesin motor domain reveals a structural similarity to myosin.
Nature 1996, 380:550-555.
Hancock WO, Howard J: Processivity of the motor protein kinesin
requires two heads. J Cell Biol 1998, 140:1395-1405.
Kozielski F, Sack S, Marx A, Thormahlen M, Schonbrunn E, Biou V,
Thompson A, Mandelkow EM, Mandelkow E: The crystal structure of
dimeric kinesin and implications for microtubule-dependent
motility. Cell 1997, 91:985-994.
10. Hoenger A, Thormahlen M, Diaz-Avalos R, Doerhoefer M,
Goldie KN, Muller J, Mandelkow E: A new look at the microtubule
binding patterns of dimeric kinesins. J Mol Biol 2000,
297:1087-1103.
11. Rice S, Lin AW, Safer D, Hart CL, Naber N, Carragher BO, Cain SM,
•
Pechatnikova E, Wilson-Kubalek EM, Whittaker M et al.: A structural
change in the kinesin motor protein that drives motility. Nature
1999, 402:778-784.
Cryo-EM studies reveal nucleotide-dependent conformational changes in
the neck linker of kinesin monomers bound to microtubules, consistent with
earlier cryo-EM data on kinesin dimers. Spectroscopic analysis of kinesin
monomers is inconsistent with existing data.
12. Mandelkow E, Hoenger A: Structures of kinesin and kinesinmicrotubule interactions. Curr Opin Cell Biol 1999, 11:34-44.
13. Fisher AJ, Smith CA, Thoden JB, Smith R, Sutoh K, Holden HM,
Rayment I: X-ray structures of the myosin motor domain of
Dictyostelium discoideum complexed with MgADP.BeFx and
MgADP.AlF4. Biochemistry 1995, 34:8960-8972.
14. Vale RD, Milligan RA: The way things move: looking under the
•
hood of molecular motor proteins. Science 2000, 288:88-95.
A provocative comparison of potential motility mechanisms in kinesin and
myosin. The authors make the astonishing hypothesis that the crystal structure of dimeric kinesin containing ADP in each head actually represents the
conformation of the molecule in the ATP-bound state.
15. Wells AL, Lin AW, Chen LQ, Safer D, Cain SM, Hasson T,
•• Carragher BO, Milligan RA, Sweeney HL: Myosin VI is an actinbased motor that moves backwards. Nature 1999,
401:505-508.
Myosin VI, which has a large insert in its converter region, moves in the
opposite direction to all other characterized myosin motors.
16. Howard J: The movement of kinesin along microtubules. Annu Rev
Physiol 1996, 58:703-729.
17.
Endow SA: Determinants of molecular motor directionality. Nat
Cell Biol 1999, 1:E163-E167.
Conformational changes during kinesin motility Schief and Howard
18. Woehlke G, Schliwa M: Directional motility of kinesin motor
proteins. Biochim Biophys Acta 2000, 1496:117-127.
19. Wade RH, Kozielski F: Structural links to kinesin directionality and
movement. Nat Struct Biol 2000, 7:456-460.
20. Stewart RJ, Thaler JP, Goldstein LS: Direction of microtubule
movement is an intrinsic property of the motor domains of
kinesin heavy chain and Drosophila ncd protein. Proc Natl Acad
Sci USA 1993, 90:5209-5213.
21. Endow SA, Higuchi H: A mutant of the motor protein kinesin that
•
moves in both directions on microtubules. Nature 2000,
406:913-916.
Mutation of a single neck residue in ncd can disrupt directional control of the
motor and cause it to move toward both the plus and minus ends of a microtubule. This finding opens up the possibility that the direction of a motor
might be regulated in cells (e.g. by phosphorylation).
22. deCastro MJ, Ho CH, Stewart RJ: Motility of dimeric ncd on a
•• metal-chelating surfactant: evidence that ncd is not processive.
Biochemistry 1999, 38:5076-5081.
The authors describe the development of a new surface-immobilization
method for histidine-tagged proteins (the surfactant F108-NTA). This paper
provides strong evidence that ncd is not processive.
23. deCastro MJ, Fondecave RM, Clarke LA, Schmidt CF, Stewart RJ:
•• Working strokes by single molecules of the kinesin-related ncd
microtubule motor. Nat Cell Biol 2000, 2:724-729.
Optical tweezers and a three-bead, suspended microtubule assay are used
to confirm that ncd is not processive. ATP-triggered working strokes of single
ncd motors at the end of ncd–microtubule binding events are resolved.
24. Mehta AD, Rock RS, Rief M, Spudich JA, Mooseker MS, Cheney RE:
•
Myosin-V is a processive actin-based motor. Nature 1999,
400:590-593.
This is the first direct evidence that myosin V moves processively on
actin filaments.
25. De La Cruz EM, Wells AL, Rosenfeld SS, Ostap EM, Sweeney HL:
•
The kinetic mechanism of myosin V. Proc Natl Acad Sci USA
1999, 96:13726-13731.
A detailed kinetic analysis of a monomeric construct of myosin V indicates
that the rate-limiting step of the actin-activated ATPase is the release of ADP.
26. Walker ML, Burgess SA, Sellers JR, Wang F, Hammer JA III,
•• Trinick J, Knight PJ: Two-headed binding of a processive myosin to
F-actin. Nature 2000, 405:804-807.
This study describes the detection of two-headed binding of Myosin V to
actin filaments by electron microscopy.
27.
•
Rief M, Rock RS, Mehta AD, Mooseker MS, Cheney RE, Spudich JA:
Myosin-V stepping kinetics: a molecular model for processivity.
Proc Natl Acad Sci USA 2000, 97:9482-9486.
The authors provide definitive proof that Myosin V is processive.
28. Sakamoto T, Amitani I, Yokota E, Ando T: Direct observation of
•
processive movement by individual myosin V molecules. Biochem
Biophys Res Commun 2000, 272:586-590.
Single fluorescently labelled myosin V molecules move several micrometers
along actin filaments before dissociating. Solution ATPase measurements in
this study may be affected by tail inactivation of the motors.
29. Howard J: Molecular motors: structural adaptations to cellular
functions. Nature 1997, 389:561-567.
30. Howard J, Hudspeth AJ, Vale RD: Movement of microtubules by
single kinesin molecules. Nature 1989, 342:154-158.
31. Crevel I, Carter N, Schliwa M, Cross R: Coupled chemical and
•
mechanical reaction steps in a processive Neurospora kinesin.
EMBO J 1999, 18:5863-5872.
This is a comprehensive biochemical and biophysical analysis of the
nucleotide-dependent interaction between Neurospora kinesin and microtubules, illustrating essential consistency with conventional kinesin. It confirms a key aspect of the hand-over-hand model of processivity — that release
of the rear head is contingent on attachment of the forward head.
32. Hoenger A, Sack S, Thormahlen M, Marx A, Muller J, Gross H,
Mandelkow E: Image reconstructions of microtubules decorated with
monomeric and dimeric kinesins: comparison with X-ray structure
and implications for motility. J Cell Biol 1998, 141:419-430.
33. Hirose K, Lowe J, Alonso M, Cross RA, Amos LA: Congruent
•
docking of dimeric kinesin and ncd into three-dimensional
electron cryomicroscopy maps of microtubule-motor ADP
complexes. Mol Biol Cell 1999, 10:2063-2074.
The authors present cryo-EM maps of kinesin bound to microtubules in the
presence of ADP. They propose a docking of kinesin·ADP onto the microtubule
27
that is consistent with proteolysis and mutagenesis studies and that avoids
steric clash between kinesin’s dimerization domain and the microtubule. They
show that the same orientation is appropriate for the docking of ncd·ADP into
a cryo-EM map of microtubules decorated with ncd in the presence of ADP.
34. Ishiwata S, Kawaguchi K: Nucleotide and loading rate-dependent
switching of single- to double-loaded binding of kinesin.
Biophys J 2000, 78:278A.
35. Okada Y, Hirokawa N: A processive single-headed motor: kinesin
•• superfamily protein KIF1A. Science 1999, 283:1152-1157.
The first report that a single-headed motor protein — a construct of
KIF1A — can move processively. The motion is accounted for by a biasedBrownian movement model. The K loop is proposed as the domain that
maintains the motor’s association with the microtubule.
36. Kikkawa M, Okada Y, Hirokawa N: 15 Å resolution model of the
•
monomeric kinesin motor, KIF1A. Cell 2000, 100:241-252.
High resolution EM density map of kinesin-related KIF1A bound to a microtubule in the presence of AMP-PNP. Docking of atomic models and gold
labelling suggest that the lysine-rich loop of KIF1A (which is absent from
kinesin) forms a novel microtubule-binding site.
37.
•
Okada Y, Hirokawa N: Mechanism of the single-headed processivity:
diffusional anchoring between the K-loop of kinesin and the C
terminus of tubulin. Proc Natl Acad Sci USA 2000, 97:640-645.
The authors characterize the nucleotide-dependent binding of the KIF1A
K-loop mutants to microtubules. Proteolysis indicates that the binding partner of the positively charged K-loop is the negatively charged carboxyl tail of
tubulin. Processivity of K-loop mutants increased with the number of positive
charges in the K-loop.
38. Pierce DW, Hom-Booher N, Otsuka AJ, Vale RD: Single-molecule
behavior of monomeric and heteromeric kinesins. Biochemistry
1999, 38:5412-5421.
39. Yonekawa Y, Harada A, Okada Y, Funakoshi T, Kanai Y, Takei Y,
Terada S, Noda T, Hirokawa N: Defect in synaptic vesicle precursor
transport and neuronal cell death in KIF1A motor protein-deficient
mice. J Cell Biol 1998, 141:431-441.
40. Cross RA, Crevel I, Carter NJ, Alonso MC, Hirose K, Amos LA: The
•
conformational cycle of kinesin. Philos Trans R Soc Lond B Biol
Sci 2000, 355:459-464.
Illuminating review of an array of data on kinesin motility, arguing for a
scheme whose key feature is the strain-sensitive trapping of ADP.
41. Schnitzer MJ, Visscher K, Block SM: Force production by single
kinesin motors. Nat Cell Biol 2000, 2:718-723.
42. Hackney DD: Evidence for alternating head catalysis by kinesin
during microtubule-stimulated ATP hydrolysis. Proc Natl Acad Sci
USA 1994, 91:6865-6869.
43. Hirose K, Lockhart A, Cross RA, Amos LA: Three-dimensional
cryoelectron microscopy of dimeric kinesin and ncd motor domains
on microtubules. Proc Natl Acad Sci USA 1996, 93:9539-9544.
44. Arnal I, Wade RH: Nucleotide-dependent conformations of the
kinesin dimer interacting with microtubules. Structure 1998,
6:33-38.
45. Coy DL, Wagenbach M, Howard J: Kinesin takes one 8-nm step
•
for each ATP that it hydrolyzes. J Biol Chem 1999,
274:3667-3671.
The authors showed directly that each ATP hydrolyzed by kinesin results
(with a high probability) in an 8 nm step.
46. Iwatani S, Iwane AH, Higuchi H, Ishii Y, Yanagida T: Mechanical and
•
chemical properties of cysteine-modified kinesin molecules.
Biochemistry 1999, 38:10318-10323.
This study provides strong confirmation of kinesin’s tight coupling using cysteine-modified motors: as the number of modified cysteines increased, the
step size remains fixed whereas the motor velocities and maximum ATPase
rates declines to the same extent.
47.
Hua W, Young EC, Fleming ML, Gelles J: Coupling of kinesin steps
to ATP hydrolysis. Nature 1997, 388:390-393.
48. Schnitzer MJ, Block SM: Kinesin hydrolyses one ATP per 8-nm
step. Nature 1997, 388:386-390.
49. Visscher K, Schnitzer MJ, Block SM: Single kinesin molecules
•
studied with a molecular force clamp. Nature 1999, 400:184-189.
An optical trap with a force-clamp provides evidence for two load-dependent
transitions, and for tight coupling under loads up to ≈5 pN.
50. Crevel IM, Lockhart A, Cross RA: Weak and strong states of kinesin
and ncd. J Mol Biol 1996, 257:66-76.
28
Cytoskeleton
51. Coy DL, Hancock WO, Wagenbach M, Howard J: Kinesin’s tail
•
domain is an inhibitory regulator of the motor domain. Nat Cell
Biol 1999, 1:288-292.
Inhibition of kinesin’s ATPase activity is relieved when artificial cargo is bound to
the tail domain. Deletion of the tail activates the ATPase without the need for
cargo binding. Inhibition is re-established by addition of exogenous tail peptide.
52. Hackney DD, Stock MF: Kinesin’s IAK tail domain inhibits initial
microtubule-stimulated ADP release. Nat Cell Biol 2000, 2:257-260.
53. Vale RD, Case R, Sablin E, Hart C, Fletterick R: Searching for
kinesin’s mechanical amplifier. Philos Trans R Soc Lond B Biol Sci
2000, 355:449-457.
54. Stone DB, Hjelm RP Jr, Mendelson RA: Solution structures of
dimeric kinesin and ncd motors. Biochemistry 1999, 38:4938-4947.
55. Ma YZ, Taylor EW: Interacting head mechanism of microtubulekinesin ATPase. J Biol Chem 1997, 272:724-730.
56. Thormahlen M, Marx A, Sack S, Mandelkow E: The coiled-coil helix
in the neck of kinesin. J Struct Biol 1998, 122:30-41.
57.
Ma YZ, Taylor EW: Kinetic mechanism of a monomeric kinesin
construct. J Biol Chem 1997, 272:717-723.
58. Young EC, Mahtani HK, Gelles J: One-headed kinesin derivatives
move by a nonprocessive, low-duty ratio mechanism unlike that
of two-headed kinesin. Biochemistry 1998, 37:3467-3479.
59. Svoboda K, Block SM: Force and velocity measured for single
kinesin molecules. Cell 1994, 77:773-784.
60. Kojima H, Muto E, Higuchi H, Yanagida T: Mechanics of single
kinesin molecules measured by optical trapping nanometry.
Biophys J 1997, 73:2012-2022.
61. Meyhöfer E, Howard J: The force generated by a single kinesin
molecule against an elastic load. Proc Natl Acad Sci USA 1995,
92:574-578.
62. Coppin CM, Pierce DW, Hsu L, Vale RD: The load dependence of
kinesin’s mechanical cycle. Proc Natl Acad Sci USA 1997,
94:8539-8544.
63. Kawaguchi K, Ishiwata S: Temperature dependence of force,
velocity, and processivity of single kinesin molecules. Biochem
Biophys Res Commun 2000, 272:895-899.
64. Keller D, Bustamante C: The mechanochemistry of molecular
motors. Biophys J 2000, 78:541-556.
65. Fisher ME, Kolomeisky AB: The force exerted by a molecular motor.
Proc Natl Acad Sci USA 1999, 96:6597-6602.
66. Ruff C, Furch M, Brenner B, Manstein DJ, Meyhoefer E: Singlemolecule tracking of myosins with genetically engineered
amplifier domains. Nat Struct Biol 2000, in press.
67.
•
Kozielski F, Svergun D, Zaccai G, Wade RH, Koch MH: The overall
conformation of conventional kinesins studied by small-angle
X-ray and neutron scattering. J Biol Chem 2000, in press.
Low angle X-ray and neutron scattering are employed to investigate the solution structures of several kinesin constructs of different lengths. For dimeric
constructs similar in length to the dimeric construct that has been crystallized [9], the scattering patterns and the radius of gyration were in good
agreement with the crystal structure.
68. Hirose K, Henningsen U, Schliwa M, Toyoshima C, Shimizu T,
Alonso M, Cross RA, Amos LA: Structural comparison of dimeric
EgS, Neurospora kinesin (Nkin) and Ncd head–Nkin neck chimera
with conventional kinesin. EMBO J 2000, 19:5308-5314.