Mole C Machines V
Mole C Machines V
Mole C Machines V
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
10.1146/annurev.bioeng.6.040803.140143
MOLECULAR MACHINES
C. Mavroidis,1 A. Dubey,2 and M.L. Yarmush3
1
CONTENTS
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ATP-BASED PROTEIN MOLECULAR MACHINES . . . . . . . . . . . . . . . . . . . . . . . .
The F0 F1 -ATP Synthase Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Kinesin, Myosin, Dynein, and Flagella Molecular Motors . . . . . . . . . . . . . . . .
DNA-BASED MOLECULAR MOTORS/DEVICES . . . . . . . . . . . . . . . . . . . . . . . . . .
The DNA Tweezers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Rotary DNA Actuator Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
INORGANIC (CHEMICAL) MOLECULAR MACHINES . . . . . . . . . . . . . . . . . . . .
The Rotaxanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Catenanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Other Inorganic Molecular Machines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
OTHER PROTEIN-BASED MOTORS UNDER DEVELOPMENT . . . . . . . . . . . . . .
Viral Protein Linear Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Synthetic Contractile Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1523-9829/04/0815-0363$14.00
364
364
365
368
375
377
377
378
378
379
381
382
382
382
383
363
13 Jul 2004
12:17
364
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
INTRODUCTION
Molecular machines can be defined as devices that can produce useful work through
the interaction of individual molecules at the molecular scale of length. A convenient unit of measurement at the molecular scale would be a nanometer. Hence,
molecular machines also fall into the category of nanomachines. Molecular machines depend on inter- and intramolecular interactions for their function. These
interactions include forces such as the ionic and Van der Waals forces and are a
function of the geometry of the individual molecules. The interaction between two
given molecules can be well understood by a set of laws governing them, which
brings in a definite level of predictability and controllability of the underlying
mechanics. Mother Nature has her own set of molecular machines that have been
working for centuries and have become optimized for performance and design
over the ages. As our knowledge and understanding of these numerous machines
continues to increase, we now see a possibility of using the natural machines, or
creating synthetic ones from scratch, by mimicking nature. In this review, we try
to understand the principles, theory, and utility of the known molecular machines
and look into the design and control issues for creation and modification of such
machines. A majority of natural molecular machines are protein based, whereas the
DNA-based molecular machines are mostly synthetic. Nature deploys proteins to
perform various cellular tasks, from moving cargo to catalyzing reactions, whereas
DNA has been retained as an information carrier. Hence, it is understandable that
most of the natural machinery is built from proteins. With the powerful crystallographic techniques now available, protein structures are clearer than ever. The
ever-increasing computing power makes it possible to dynamically model protein
folding processes and predict the conformations and structure of lesser known
proteins. These findings help unravel the mysteries associated with the molecular
machinery and pave the way for the production and application of these miniature
machines in various fields, including medicine, space exploration, electronics and
military. We divide the molecular machines into three broad categoriesprotein
based, DNA-based, and chemical molecular motors.
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
P1: IKH
365
functions both inside and outside cells for millions of years, have now been segregated out of their natural environment and are seen as energy conversion devices
to obtain forces, torques, and motion. One disadvantage associated with ATP dependence is that the ATP creation machinery itself could be many times heavier
and bulkier than the motors, thereby making the assembly more complex. These
machines perform best in their natural environment, and in the near future it may
not be possible to have them as a part of feasible biomimetic molecular machinery.
13 Jul 2004
12:17
366
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
Figure 1 The F0F1-ATPase motors. The F0 motor is embedded in the inner mitochondrial membrane of the mitochondria. F0 is typically composed of a, b, and c subunits
as shown. The F1 motor is the soluble region composed of three -, three -, one each
of -, - and -subunits.
alternately so that they make a symmetric circular pattern when viewed from
the top. There are -subunits attached to the periphery of the - cylinder and the
-subunits are present at the base of the -subunits, as shown in Figure 1. Hence,
the F1 motor is composed of nine polypeptides (10). The - and the -subunits
contain nucleotide-binding sites that bind ATP/ADP molecules. The nucleotidebinding sites in the -subunits simply bind the nucleotide, whereas those in the
-subunit actually perform the catalysis. The a, b, and the c subunits shown in
Figure 1 are a part of the F0 motor discussed below.
The binding-change mechanism to explain the function of F1-ATPase was proposed in 1973 (11). The mechanism, as known today,
shows that each of the -subunits take three forms: O (open), L (loose), and
T (tight) binding site. When the subunit is in the O form, it is catalytically inactive
and has very low affinity to bind substrates. In the L form, the subunit loosely binds
substrates (ADP and Pi), although it is catalytically inactive. In the T form, the
ADP and Pi are converted into a tightly bound ATP until a conformational change
converts the T-site into an O-site, thereby allowing the release of the newly formed
ATP (12). The mechanism is shown in Figure 2.
The conformational change in the -subunits is triggered by the rotation of
the 4.5 nm long -subunit, which acts as a link connecting the F1-ATPase to the
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
MOLECULAR MACHINES
367
F0-ATPase. This was shown experimentally in 1997 (13). In this experiment, the F1ATPase was attached to a nickel-coated glass surface; a 13 m long fluorescently
labeled actin filament was attached to the other end of the -subunit. The rotation
could then be observed through a fluorescence microscope, which was extremely
interesting because the motor has a diameter of about 10 nm, whereas it could
support and rotate a structure about one hundred times larger! However, the rate of
rotation was reduced by 50 times to 1 rotation per second. The experimental setup
is shown in Figure 3. F1-ATPase can produce 80100 pN-nm of rotary torque (4).
Since the rotation of the -subunit has been shown
to play the essential role in ATP creation, it is now imperative to see what causes the
13 Jul 2004
12:17
368
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
-subunit to rotate. The answer lies in the functioning of the F0-ATPase motor. Although the structure of the F0-ATPase is not as well known as that of the F1-ATPase,
findings indicate that its structure depends on its source. The F0 domain from the
eubacterial enzymes, exemplified by E. coli. (14), has three types of subunits
termed a, b, and c with 1 unit of a, 2 units of b, and 9 to 12 (15) units of c subunits
(16, 17). Hence the subunit a, the two b, and the twelve c subunits in Figure 1
belong to the F0-ATPase motor. In yeast, the F0 has only 10 c subunits (18). The F0
turbine from plant chloroplasts was found to have 14 c subunits (19), whereas
a Na+-driven specimen from bacteria was found to have 14 such subunits (20).
A flow of ions through the membrane propels the
observed reversible (21) rotation of F0-ATPase (22). As mentioned above, F0 is
the membrane-spanning unit of the ATPase motor. It remains embedded in the
mitochondrial or cellular membrane. In 1978, it was discovered that a chemical
potential gradient for protons is formed across the inner mitochondrial membrane
or the proton-motive force (8). This force is utilized by the ATPase enzyme to
produce ATP. The structure of F0-ATPase is not as well known as its F1 counterpart,
which has been fully resolved (18, 2328). The mechanochemical and quantitative
models that explain how the ion-motive force is converted into the rotation of the
-subunit were described in (15, 2933).
The first hybrid nanoassembly structures powered by F1-ATPase was proposed
in (34). Nano-fabricated Ni posts, about 80 nm in diameter and 200 nm in height,
each separated by about 2.5 m were built. Upon these posts they attached specially
produced recombinant biotinylated F1-ATPases using histidine tags into their subunit coding sequences. A streptavidin molecule was bound to the -subunit,
and finally Ni propellers of lengths 750 to 1400 nm were attached to them. In
an action that is reverse of its ATP-producing cycle, the F1-ATPase consumed
externally provided ATP and produced anticlockwise rotation with a speed of
about eight rotations per second. To date, this achievement remains a landmark in
bionanotechnology.
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
P1: IKH
369
only part they have in common is the catalytic portion known as the motor domain.
They have significant differences in their location within cells, their structural
organization, and the movement they generate (37). Muscle myosin, whose study
dates back to 1864, has served as a model system for understanding motility for
decades. Kinesin, however was not discovered until 1985, using in vitro motility
assays (38). Conventional kinesin is a highly processive motor that can take several
hundred steps on a microtubule without detaching (39, 40), whereas muscle myosin
executes a single stroke and then dissociates (41). Detailed analysis and modeling
of these motors has been done (38, 42).
Kinesin and myosin make for an interesting comparison. Kinesin is microtubule
based; it binds to and carries cargoes along microtubules, whereas myosin is actin
based. The motor domain of kinesin weighs one third that of myosin and one
tenth of that of dynein (43). Before the advent of modern microscopic and analytic
techniques, it was believed that these two had little in common. However, the
crystal structures available today indicate that they probably originated from a
common ancestor (44).
Myosin is a diverse superfamily of motor proteins
(45). Myosin-based molecular machines transport cargoes along actin filaments,
the two-stranded helical polymers of the protein actin that are about 59 nm in
diameter. They do this by hydrolyzing ATP and utilizing the released energy (46).
In addition to transport, they are also involved in the process of force generation during muscle contraction, wherein thin actin filaments and thick myosin
filaments slide past each other. Not all members of the myosin superfamily have
been characterized. However, much is known about their structure and function.
Myosin molecules were first seen (in the late 1950s) through electron microscope
protruding out from thick filaments and interacting with the thin actin filaments
(4749). Since that time, ATP has been known to play a role in myosin-related
muscle movement along actin (50); however, the exact mechanism was unknown,
until it was explained in (51).
13 Jul 2004
12:17
370
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
The motor domain in itself is sufficient for moving actin filaments (58). Threedimensional structures of a myosin head revealed that it is a pear-shaped domain,
about 19 nm long and 5 nm in maximum diameter (58, 59).
A crossbridge-cycle model for the action of myosin on actin has been widely accepted since 1957 (47, 60, 61). Since
the time the atomic structures of actin monomer (62, 63) and myosin (59) were
resolved, this model has been refined into a lever-arm model which is now acceptable (64). Only one motor head is able to connect to the actin filament at a time,
the other head remains passive. Initially, the catalytic domain in the head has ADP
and Pi bound to it and, as a result, its binding with actin was weak. With the active
motor head docking properly to the actin-binding site, the Pi has to be released.
As soon as this happens, the lever arm swings counterclockwise (65) owing to
a conformational change (49, 6671), which pushes the actin filament down by
about 10 nm along its longitudinal axis (38). The active motor head now releases
its bound ADP, and another ATP molecule, by way of Brownian motion, quickly
replaces it, making the binding of the head to the actin filament weak again. The
myosin motor then dissociates from the actin filament, and a new cycle starts.
However, nano-manipulation of single S1 molecules (motor domains) shows that
myosin can take multiple steps per ATP molecule hydrolyzed, moving in 5.3 nm
steps and resulting in displacements of 11 to 30 nm (72).
Much structural information about kinesin is now available through the crystal structures (44, 78, 79). The motor domain
contains a folding motif similar to that of myosin and G proteins (36). The two
heads or the motor domains of kinesin are linked via neck linkers to a long coiled
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
P1: IKH
371
coil, which extends up to the cargo (Figure 4b). These heads interact with the and -subunits of the tubulin hetrodimer along the microtubule protofilament. The
heads contain nucleotide- and microtubule-binding domains.
Although kinesin is also a two-headed
linear motor, its modus operandi is different from myosin in the sense that both
of its heads work together in a coordinated manner in contrast to one being left
out in the case of myosin. Figure 4b shows the kinesin walk. Each of the motor
heads is near the microtubule in the initial state, with each motor head carrying an
ADP molecule. When one of the heads loosely binds to the microtubule, it loses
its ADP molecule to facilitate a stronger binding. Another ATP molecule replaces
the ADP, which facilitates a conformational change such that the neck region
of the bound head snaps forward and zips on to the head (37). In the process, it
pulls the other ADP-carrying motor head forward by about 16 nm so that it can bind
to the next microtubule-binding site. This results in the net movement of the cargo
by about 8 nm (80). The second head now binds to the microtubule by losing its
ADP, which is promptly replaced by another ATP molecule (Brownian motion).
The first head, meanwhile hydrolyzes the ATP and loses the resulting Pi. It is
then snapped forward by the second head while it carries its ADP forward. Hence
coordinated hydrolysis of ATP in the two motor heads is the key to the kinesin
processivity (81, 82). Kinesin is able to take about 100 steps before detaching from
the microtubule (39, 76, 83), while moving at 1000 nm/s and exerting forces of
the order of 56 pN at rest (84, 85).
13 Jul 2004
12:17
372
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
Figure 5 A dynein molecule. Shown are the globular heads (heavy chains) connected
to the intermediate chains and the light chains. Dynactin complex components p150,
p135, dynamitin, p62, capping proteins, Arp1, and Actin are also shown.
(104106), which indicate that its mode of working will be entirely different from
kinesins and myosins. This puts dyneins into the AAA superfamily of mechanoenzymes. The dynein heavy chains contain six tandemly linked AAA modules (107,
108), with the head having a ring-like domain organization, which is typical of a
AAA superfamily. Four of these are nucleotide-binding motifs, named P1P4, but
only P1 (AAA1) is able to hydrolyze ATP.
Because dynein is a larger and more
complex structure than other motor proteins, its mode of operation is not as well
known. However, electron microscopy and image processing was used (100) to
show the structure of a flagellar dynein at the start and end of its power stroke, which
gives some insight into its possible mode of force generation. When the dynein
contains bound ADP and Vi (vandate), it is in the prepower stroke conformation.
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
P1: IKH
373
The state when it has lost the two, known as the apo-state, is the more compact
post-power stroke state. There is a distinct conformational change involving the
stem, linker, head, and the stalk that produces about 15 nm of translation onto the
microtubule bound to the stalk (100).
Unicellular organisms such as E. coli have an interesting
mode of motility (see 109111). They have a number of molecular motors, about
45 nm in diameter, that drive their feet or the flagella, which help the cell to swim.
Motility is critical for cells, as they often have to travel from a less favorable to
a more favorable environment. The flagella are helical filaments that extend out
of the cell into the medium and perform a function analogous to what the oars
perform to a boat. The flagella and the motor assembly are called a flagellum. The
motor assembly imparts a rotary motion into the flagella (112, 113). In addition
to a rotary mechanism, the flagellar machines consist of components such as rate
meters, particle counters, and gearboxes (114). These are necessary to help the cell
decide which way to go, depending on the change of concentration of nutrients in
the surroundings. The rotary motion imparted to the flagella needs to be modulated
to ensure the cell is moving in the proper direction, as well as to ensure that all
flagella of the given cell are providing a concerted effort toward it (115). When
the motors rotate the flagella in a counterclockwise direction, as viewed along the
flagella filament from outside, the helical flagella create a wave away from the cell
body. Adjacent flagella subsequently intertwine in a propulsive corkscrew manner
and propel the bacteria. When the motors rotate clockwise, the flagella fly apart,
causing the bacteria to tumble or change its direction (116). These reversals occur
irregularly, giving the bacterium a random walk, unless, of course, there is a preferential direction of motility due to reasons mentioned earlier. The flagella motors
allow the bacteria to move at speeds of as much as 25 m/s, with directional reversals occurring approximately 1 per second (117). A number of bacterial species in
addition to E. coli depend on flagella motors for motility: e.g., Salmonella enterica serovar, Typhimurium (Salmonella), Streptococcus, Vibrio spp., Caulobacter,
Leptospira, Aquaspirrilum serpens, and Bacillus. The rotation of flagella motors
is stimulated by a flow of ions through them, which is a result of a build-up of
a transmembrane ion gradient. There is no direct ATP-involvement; however, the
proton gradient needed for the functioning of flagella motors can be produced by
ATPase.
13 Jul 2004
12:17
374
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
FlhAFlhE; 19 Fli-type proteins, FliAFliT; with MotA and MotB making a total
of 40 related proteins. The group names Flg, Flh, Fli, and Mot correspond to the
related genes (120). Within the main structural proteins are other proteins: FliC
or the filament; FliD (filament cap); FliF or the MS-ring; FliG, FliM, and FliN
(C-ring); FlgB, FlgC, and FlgF (proximal rod); FlgG (distal rod); FlgH (L-ring);
FlgI (P-ring); FlgK and FlgL (hook-filament junction); and MotA-MotB (torquegenerating units) (see Figure 6). It was initially believed that the M and S were
two separate rings (M, membrane; S, supramembranous) (121). However, they
are now called the MS-ring because they were found to be two domains of the
same protein, FliF (122, 123). The C-ring stands for cytoplasmic (124126); the
e names for the P and L-rings come from peptidoglycan and lipopolysaccharide,
respectively, indicating their location as seen in Figure 6. FlhA, B, FliH, I, O, P,
Q, and R constitute the transport apparatus.
The hook and filament part of the flagellum is located outside the cell body. The
motor portion is embedded in the cell membrane, with the C-ring and the transport
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
P1: IKH
375
apparatus inside the inner membrane in the cytoplasmic region. MotA and MotB
are arranged in a circular array embedded in the inner membrane, with the MS-ring
at the center. Connected to the MS-ring is the proximal end of a shaft, to which the
P-ring, embedded in the peptidoglycan layer, is attached. Moving further outward,
is the L-ring, which is embedded in the outer cell membrane, followed by the distal
shaft end that protrudes out of the cell. To this end there is an attachment of the
hook and the filament, both of which are polymers of hook-protein and flagellin
respectively.
Function: the flagella motors The flagellar motors in most cases are powered by
protons flowing through the cell membrane (proton-motive force) barring exceptions such as certain marine bacteria, for example, the Vibrio spp., which are driven
by Na+ ions (127). There are about 1200 protons required to rotate the motor by
one rotation (128). A complete explanation of how this proton flow is able to generate torque is not yet available. From what is known, the stator units of MotA
and MotB play an important role in torque generation. They form a MotA/MotB
complex that when oriented properly binds to the peptidoglycan and opens proton
channels through which protons can flow (129). It is believed that there are eight
such channels per motor (130). The proton-motive force is a result of the difference
of pH between the outside and inside of the cell. The E. coli cells like to maintain an internal pH of 7.67.8, so depending on the pH of the surroundings, the
proton-motive force will vary, and hence the speed of rotation of their motors. To
test how the speed of rotation depends on the proton-motive force, the motors were
powered by external voltage with attached markers acting as heavy loads (131).
As expected, the rotation was found to depend directly on the proton-motive force.
According to the most widely accepted model, MotA/MotB complex interacts with
the rotor via binding sites. The passage of protons through a MotA/MotB complex
(stator or torque generator) moves it so that the protons bind to the next available
binding site on the rotor, thereby stretching their linkage. When the linkage recoils,
the rotor assembly has to rotate by one step. Hence whichever complex receives
protons from the flux will rotate the rotor and generate torque. The torque-speed
dependence of the motor has been studied in detail (132, 133) and indicates the
torque range of about 2700 to 4600 pN-nm.
13 Jul 2004
12:17
376
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
predictable behavior. There are certain qualities that make DNA an attractive choice
for the construction of artificial nanomachines. In recent years, DNA has found
use not only in mechanochemical but also in nanoelectronic systems (134137).
A DNA double-helical molecule is about 2 nm in diameter and has 3.43.6-nm
helical pitch no matter what its base composition is, a structural uniformity not
achievable with protein structures if one changes their sequence. Furthermore,
double-stranded DNA (ds-DNA) has a respectable persistence length of about
50 nm (138), which provides it enough rigidity to be a candidate component of
molecular machinery. Single-stranded DNA (ss-DNA) is very flexible and cannot
be used where rigidity is required; however, this flexibility allows its application in
machine components such as hinges or nanoactuators (139). Its persistence length
is about 1 nm, covering up to 3 base pairs (140) at 1M salt concentration.
Other than the above structural features, two important and exclusive properties
make DNA suitable for molecular level constructions: molecular recognition and
self-assembly. The nucleotide bases A and T on two different ss-DNA have affinity
for each other, so do G and C. Effective and stable ds-DNA structures are formed
only if the base orders of the individual strands are complementary. Hence, if
two complementary single strands of DNA are in a solution, they will eventually
recognize each other and hybridize, or zip-up, forming a ds-DNA. This property of
molecular recognition and self-assembly has been exploited in a number of ways
to build complex molecular structures (141148). From a mechanical perspective,
if the free energy released by hybridization of two complementary DNA strands
is used to lift a hypothetical load, a force capacity of 15 pN can be achieved (F.C.
Simmel & B.Yurke, unpublished data), comparable to that of other molecular
machines such as kinesin (5 pN) (150).
The first artificial DNA-based structure in the form of a cube in 1991 was
presented in (143, 151). More complex structures such as knots (152, 153) and
Borromean rings (147) were also developed. In addition to these individual constructs, two-dimensional arrays (145, 154, 155) were made with the help of the
double-crossover (DX) DNA molecule (156158). This DX molecule gave the
structural rigidity required to create a dynamic molecular device, the B-Z switch
(159). DNA double helices can be of three types: A-, B-, or Z-DNA. The B-DNA is
the natural, right-handed helical form of DNA, whereas the A-DNA is a shrunken,
low-humidity form of the B-DNA. Z-DNA, obtained from certain CG base repeat sequences occurring in B-DNA, can take a left-handed double helical form
(160). The CG-repeated base pair regions can be switched between the left and the
right-handed conformations by changing ionic concentration (161). The switch
was designed in such a way that it had three cyclic strands of DNA, two of them
wrapped around a central strand that had the CG repeat region in the middle. On
the two free ends of the side strands fluorescent dyes were attached in order to
monitor the conformational change. With the change in ionic concentration the
central CG repeat sequence could alternate between the B and the Z modes bidirectionally, which was observed through fluorescence resonance energy transfer
(FRET) spectroscopy.
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
P1: IKH
377
13 Jul 2004
12:17
378
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
in a linear array (166). Because of the larger size of the structures, they could be
visualized using an atomic force microscope to prove that the rotary device indeed
rotates. A possible application of two DNA rotary machines to rotate a central disc
is shown in Figure 9.
The Rotaxanes
Rotaxane family of molecular machines is characterized by two parts: a dumbbellshaped compound with two heavy chemical groups at the ends and a light, cyclic
component, called a macrocycle, interlocked between the heads (Figure 10).
A reversible switch can be made with a rotaxane setup (172). For this, one needs
to have two chemically active recognition sites in the neck region of the dumbbell.
In this particular example, the thread was made of polyether, marked by recognition sites of hydroquinol units and terminated at the ends by large triisoproplylsilyl
groups. A tetracationic bead was designed and self-assembled into the system that
interacts with the recognition sites. The macrocycle has a natural, low-energy state
on the first recognition site, but can be switched reversibly between the two sites
upon application of suitable stimuli. Depending on the type of rotaxane setup,
the stimuli can be chemical, electrochemical, or photochemical (173, 174). The
stereo-electronic properties of the recognition sites can be altered by protonation
or deprotonation, or by oxidation or reduction, thereby changing the affinity of the
sites toward the macrocycle. In a recent example, light-induced acceleration of rotaxane motion was achieved by photoisomerization (175). Similar controls through
alternating current (oscillating electric fields) had previously been shown (176).
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
P1: IKH
379
Figure 10 A typical rotaxane shuttle setup. The macrocycle encircles the thread-like
portion of the dumbbell with heavy groups at its ends. The thread has two recognition
sites that can be altered reversibly so as to make the macrocycle shuttle between the
two sites.
There are various ways for making rotaxanes by supramolecular synthesis (177).
They can be self-assembled (178) using template-directed synthesis (179) methods
such as threading, clipping, and slippage (180182). In addition, various other rotaxane shuttles and means of controlling the switching motion have been described
(183192).
The Catenanes
The catenanes are also a special type of interlocked structures that represent a
growing family of molecular machines. They are synthesized by supramolecular
assistance to molecular synthesis (177). The general structure of a catenane is
that of two interlocked ring-like components that are noncovalently linked via a
mechanical bond, i.e., they are held together without any valence forces. Both
macrocyclic components have recognition sites composed of atoms or groups of
atoms that are redox active or photochemically reactive. It is possible to have both
rings with similar recognition sites. In such a scenario, one of the rings may rotate
inside the other with the conformations stabilized by noncovalent interactions, but
the two states of the inner ring, differing by 180 , will be undistinguishable (degenerate) (193). For better control and distinguishable molecular conformations,
it is desirable to have different recognition sites within the macrocycles. Then they
can be controlled independently through their own specific stimuli. The stereoelectronic property of a recognition site within a macrocycle can be varied such
that at one point it has more affinity for the sites on the other ring. At this instant,
the force balance will guide the rotating macrocycle for a stable conformation,
which requires that particular site to be inside the other macrocycle. Similarly,
with other stimuli, this affinity can be turned off, or even reversed, along with an
13 Jul 2004
12:17
380
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
Figure 11 A nondegenerate catenane. One of the rings (the moving ring) has two
different recognition sites in it. Both sites can be turned off or on with different stimuli.
When the trapezoidal-shaped site is activated, the force and energy balance results
in the first conformation, whereas when the disc-shaped site is activated, the second
conformation results. They can be called states 0 and 1, analogous to binary machine
language.
increase of the affinity of the second recognition site on the rotating macrocycle
toward those on the static one. There is a need for computational modeling, simulation, and analysis of such molecular machine motion (194). Catenanes can also
be designed for chemical, photochemical, or electrochemical control (195199).
Figure 11 describes one such catenane molecular motor.
For both rotaxane- and catenane-based molecular machines, it is desirable to
have recognition sites such that they can be easily controlled externally. Hence, it is
preferable to build sites that are either redox active or photo active (173). Catenanes
can also be self-assembled (200). An example of a catenane-assembled molecular
motor is the electronically controllable bistable switch (201). An intuitive way of
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
P1: IKH
381
13 Jul 2004
12:17
382
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
along with eight other motors from the same material is now operational (229).
This redesigned motor has distinct upper and lower portions, and it operates at
a higher speed. It also provides a good example of how controlled motion at the
molecular level can be used to produce a macroscopic change in a system that
is visible to the naked eye. The light-driven motors when inside liquid crystal
(LC) films can produce a color change by inducing a reorganization of mesogenic
molecules (230).
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
P1: IKH
383
CONCLUSIONS
The recent explosion of research in nanotechnology, combined with important discoveries in molecular biology, has created a new interest in biomolecular machines
and robots. The main goal in the field of biomolecular machines is to use various
biological elementswhose function at the cellular level creates a motion, force,
or a signalas machine components that perform the same function in response
to the same biological stimuli but in an artificial setting. In this way, proteins and
DNA could act as motors, mechanical joints, transmission elements, or sensors. If
all these components were assembled together they could form nanodevices with
Figure 12 (a) VPL motor at neutral pH. Front view of the partially -helical triple
stranded coiled coil. VPL motor is in the closed conformation. (b) VPL Motor in the
open conformation at acidic pH. The random coil regions are converted into welldefined helices and an extension occurs at lower pH.
13 Jul 2004
12:17
384
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
Figure 12
AR220-BE06-15.sgm
YARMUSH
(Continued)
LaTeX2e(2002/01/18)
P1: IKH
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
P1: IKH
385
multiple degrees of freedom, able to apply forces and manipulate objects in the
nanoscale world, transfer information from the nano- to the macroscale world, and
even travel in a nanoscale environment.
The future of molecular machinery is bright. We are at the dawn of a new era
in which many disciplines will merge, including robotics, mechanical, chemical,
and biomedical engineering, chemistry, biology, physics, and mathematics, so that
fully functional systems will be developed. However, challenges toward such a
goal abound. Developing a complete database of different biomolecular machine
components and the ability to interface or assemble different machine components
are some of the challenges to be faced in the near future. The problems involved
in controlling and coordinating several biomolecular machines will come next.
ACKNOWLEDGMENTS
This work was supported by the National Science Foundation (DMI-02,28103
and DMI-03,03950). Any opinions, findings, conclusions, or recommendations
expressed in this publication are those of the authors and do not necessarily reflect
the views of the National Science Foundation. The authors thank Kevin Nikitczuk
of the Department of Biomedical Engineering at Rutgers University for providing
assistance in the creation of the graphics for this paper.
The Annual Review of Biomedical Engineering is online at
http://bioeng.annualreviews.org
LITERATURE CITED
1. Oster G. 2003. How protein motors
convert chemical energy into mechanical work. In Molecular Motors, ed. M
Schliwa, pp. 20728. New York: Wiley
2. Boyer PD. 1998. Energy, life, and ATP.
Biosci. Rep. 18:97117
3. Lipmann F. 1941. Metabolic generation
and utilization of phosphate bond energy.
Adv. Enzymol. 1:99
4. Yasuda R, Noji H, Kinosita K Jr, Yoshida
M. 1998. F1-ATPase is a highly efficient
molecular motor that rotates with discrete
120 degree steps. Cell 93:111724
5. Fernandez-Moran H. 1962. Molecular organization of cell membranes. Circulation
26:103965
6. Kagawa Y, Racker E. 1966. Partial resolution of the enzymes catalyzing oxidative phosphorylation. J. Biol. Chem.
241:247582
13 Jul 2004
12:17
386
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
Walker JE, Ueda T, et al. 1997. The crystal structure of the nucleotide-free alpha 3
beta 3 subcomplex of F1-ATPase from the
thermophilic Bacillus PS3 is a symmetric
trimer. Structure 5:82536
Bianchet MA, Hullihen J, Pedersen PL,
Amzel LM. 1998. The 2.8-A structure
of rat liver F1-ATPase: configuration of
a critical intermediate in ATP synthesis/hydrolysis. Proc. Natl. Acad. Sci. USA
95:1106570
Menz RI, Walker JE, Leslie AG. 2001.
Structure of bovine mitochondrial F1ATPase with nucleotide bound to all three
catalytic sites: implications for the mechanism of rotary catalysis. Cell 106:331
41
Groth G. 2002. Structure of spinach
chloroplast F1-ATPase complexed with
the phytopathogenic inhibitor tentoxin.
Proc. Natl. Acad. Sci. USA 99:346468
Rodgers AJ, Wilce MC. 2000. Structure of
the gamma-epsilon complex of ATP synthase. Nat. Struct. Biol. 7:105154
Dimroth P, Kaim G, Matthey U. 2000.
Crucial role of the membrane potential for
ATP synthesis by F1F0 ATP synthases. J.
Exp. Biol. 203 Pt 1:5159
Oster G, Wang H, Grabe M. 2000. How
Fo-ATPase generates rotary torque. Philos. Trans. R. Soc. London Ser. B. 355:
52328
Dimroth P. 2000. Operation of the F0 motor of the ATP synthase. Biochim. Biophys. Acta 1458:37486
Grabe M, Wang H, Oster G. 2000. The
mechanochemistry of V-ATPase proton
pumps. Biophys. J. 78:2798813
Oster G, Wang H. 2000. Reverse engineering a protein: the mechanochemistry
of ATP synthase. Biochim. Biophys. Acta
1458:482510
Soong RK, Bachand GD, Neves HP,
Olkhovets AG, Craighead HG, Montemagno CD. 2000. Powering an inorganic
nanodevice with a biomolecular motor.
Science 290:155558
Howard J. 1997. Molecular motors:
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
P1: IKH
387
13 Jul 2004
12:17
388
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
P1: IKH
389
13 Jul 2004
12:17
390
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
136. Robinson BH, Seeman NC. 1987. The
design of a biochip: a self-assembling
molecular-scale memory device. Protein
Eng. 1:295300
137. Seeman NC, Belcher AM. 2002. Emulating biology: building nanostructures from
the bottom up. Proc. Natl. Acad. Sci. USA
99:645155
138. Smith SB, Cui Y, Bustamante C. 1996.
Overstretching B-DNA: the elastic response of individual double-stranded and
single-stranded DNA molecules. Science
271:79599
139. Simmel FC, Yurke B. 2001. Using DNA to
construct and power a nanoactuator. Phys.
Rev. E. 63:041913
140. Tinland B, Pluen A, Sturm J, Weill G.
1997. Persistence length of singlestranded DNA. Macromolecules 30:
576365
141. Seeman NC. 1997. DNA components for
molecular architecture. Acc. Chem. Res.
30:34791
142. Niemeyer CM. 1999. Progress in engineering up nanotechnology devices utilizing DNA as a construction material.
Appl.Phys. A 68:11924
143. Chen JH, Seeman NC. 1991. Synthesis
from DNA of a molecule with the connectivity of a cube. Nature 350:63133
144. Zhang Y, Seeman NC. 1994. Construction
of a DNA-truncated octahedron. J. Am.
Chem. Soc. 116:166169
145. Winfree E, Liu F, Wenzler LA, Seeman
NC. 1998. Design and self-assembly of
two-dimensional DNA crystals. Nature
394:53944
146. Yan H, Zhang X, Shen Z, Seeman NC.
2002. A robust DNA mechanical device
controlled by hybridization topology. Nature 415:6265
147. Mao C, Sun W, Seeman NC. 1997. Assembly of Borromean rings from DNA.
Nature 386:13738
148. Mao C, LaBean TH, Relf JH, Seeman
NC. 2000. Logical computation using algorithmic self-assembly of DNA triplecrossover molecules. Nature 407:49396
P1: IKH
391
13 Jul 2004
12:17
392
163.
164.
165.
166.
167.
168.
169.
170.
171.
172.
173.
174.
175.
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
176.
177.
178.
179.
180.
181.
182.
183.
184.
185.
Altieri A, et al. 2003. Photoisomerization of a rotaxane hydrogen bonding template: light-induced acceleration of a large
amplitude rotational motion. Proc. Natl.
Acad. Sci. USA 100:1014
Bermudez VV, Capron N, Gase T, Gatti
FG, Kajzar F, et al. 2000. Influencing intramolecular motion with an alternating
electric field. Nature 406:60811
Fyfe MCT, Stoddart JF. 1997. Synthetic
supramolecular chemistry. Acc. Chem.
Res. 30:393401
Whitesides GM, Mathias JP, Seto CT.
1991. Molecular self-assembly and
nanochemistry: a chemical strategy for
the synthesis of nanostructures. Science
254:131219
Gavina P, Sauvage JP. 1997. Transitionmetal template synthesis of a rotaxane
incorporating two different coordinating
units in its thread. Tetrahedron Lett.
38:352124
Raymo FM, Houk KM, Stoddart JF.
1998. The mechanism of the slippage approach to rotaxanes. Origin of the all-ornothing substituent effect. J. Am. Chem.
Soc. 120:931822
Whitesides GM, Simanek EE, Mathias JP,
Seto CT, Chin DN, et al. 1995. Noncovalent synthesis: using physical-organic
chemistry to make aggregates. Acc. Chem.
Res. 28:3744
Asakawa M, Ashton PR, Ballardini R,
Balzani V, Belohradsky M, et al. 1997.
The slipping approach to self-assembling
[n]rotaxanes. J. Am. Chem. Soc. 119:302
10
Brouwer AM, Frochot C, Gatti FG, Leigh
DA, Mottier L, et al. 2001. Photoinduction
of fast, reversible translational motion in a
hydrogen-bonded molecular shuttle. Science 291:212428
Bissell RA, Cordova E, Kaifer AE, Stoddart JF. 1994. A chemically and electrochemically switchable molecular shuttle.
Nature 369:13337
Martnez-Daz M-V, Spencer N, Stoddart JF. 1997. The self-assembly of a
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
186.
187.
188.
189.
190.
191.
192.
193.
194.
195.
196.
197.
198.
199.
200.
201.
202.
203.
204.
205.
206.
P1: IKH
393
13 Jul 2004
12:17
394
207.
208.
209.
210.
211.
212.
213.
214.
215.
216.
AR
AR220-BE06-15.tex
MAVROIDIS
DUBEY
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
P1: IKH
YARMUSH
217.
218.
219.
220.
221.
222.
223.
224.
225.
226.
227.
228.
229.
13 Jul 2004
12:17
AR
AR220-BE06-15.tex
AR220-BE06-15.sgm
LaTeX2e(2002/01/18)
MOLECULAR MACHINES
230.
231.
232.
233.
near-perfect photoequilibria and acceleration of the speed of rotation by structural modification. J. Am. Chem. Soc. 124:
503751
van Delden RA, Koumura N, Harada N,
Feringa BL. 2002. Unidirectional rotary
motion in a liquid crystalline environment: color tuning by a molecular motor. Proc. Natl. Acad. Sci. USA 99:4945
49
Dubey A, Thornton A, Nikitczuk KP,
Mavroidis C, Yarmush ML. 2003. Viral
protein linear (VPL) nano-actuators. Presented at Proc. 2003 3rd IEEE Conf. Nanotechnology, San Francisco
Wilson IA, Skehel JJ, Wiley DC. 1981.
Structure of the haemagglutinin membrane glycoprotein of influenza virus at
3 A resolution. Nature 289:36673
Chan DC, Fass D, Berger JM, Kim PS.
234.
235.
236.
237.
238.
P1: IKH
395
HI-RES-BE06-15-Mavroidis.qxd
7/13/04
9:10 PM
Page 1
MOLECULAR MACHINES
C-1
Figure 4 The kinesin-myosin walks. (a) Myosin motor mechanism. (i) Motor head
loosely docking to the actin-binding site; (ii) the binding becomes tighter along with
the release of Pi; (iii) lever arm swings to the left with the release of ADP; and
(iv) replacement of the lost ADP with a fresh ATP molecule results in dissociation of
the head. (b) Kinesin heads working in conjunction. (i) Both ADP-carrying heads
come near the microtubule and one (black neck) binds; (ii) loss of bound ADP and
addition of fresh ATP in the bound head moves the other (red neck) to the right;
(iii) the second head (red) binds to microtubule while losing its ADP, and replacing
it with a new ATP molecule, whereas the first head hydrolyzes its ATP and loses Pi;
(iv) the ADP-carrying black neck will now be snapped forward, and the cycle will
be repeated.
HI-RES-BE06-15-Mavroidis.qxd
C-2
MAVROIDIS
7/13/04
DUBEY
9:10 PM
Page 2
YARMUSH
Figure 7 The DNA tweezers. (a) The machine is in the open conformation with the
central strand (black) partially hybridized to the two side strands (red and green);
(b) a fuel strand (white) is introduced; and (c) the fuel strand hybridizes with the two
free ends to bring the device into a closed conformation. With the addition of complement to the fuel strand, it can be removed to leave the system back to state (a).
Figure reprinted with permission from Bernard Yurke, Lucent Technologies.
7/13/04
9:10 PM
Figure 8 (a) PX and JX2 topological DNA motifs. The two lower double helices, C and D, are rotated by 180 during transition from PX
to JX2. (b) Working principle of the rotary DNA machine: (stage i) Type 1 strands (blue) are removable from PX motifs by the addition of
their respective complementary strands (black dots at ends); (stage ii) the addition of Type 2 strands (purple) results in a rotation of C and
D units; (stage iii) Type 2 strands can then be removed by adding their complementary strands; (stage iv) Type 1 strands added again to
revert the motif back to PX state.
HI-RES-BE06-15-Mavroidis.qxd
Page 3
MOLECULAR MACHINES
C-3
HI-RES-BE06-15-Mavroidis.qxd
C-4
MAVROIDIS
7/13/04
DUBEY
9:10 PM
Page 4
YARMUSH
Figure 9 A possible application of two DNA rotary machines to rotate a central disc
by half a rotation at a time to achieve one full rotation. Red tags on the disc are stoppers that will allow linkage to the motifs.