Essays on Science and SocietyGE PRIZE WINNER

How Molecular Motors Move

See allHide authors and affiliations

Science  10 Feb 2006:
Vol. 311, Issue 5762, pp. 792-793
DOI: 10.1126/science.1125068a

Myosin and kinesin motor proteins use the energy obtained from adenosine triphosphate (ATP) hydrolysis to transport organelles and vesicles by moving along the cytoskeleton. Structurally, these motors are dimeric, having two motor heads, two legs, and a common stalk. The head regions bind to actin or microtubule filaments and power the forward movement. The central question was how the two heads are coupled so that the motor can processively move along its track. In the hand-over-hand model (1), ATP binding and hydrolysis creates a conformational change in the forward head (head 1) and this conformation pulls the rear head (head 2) forward, while head 1 stays fixed on the track. In the next step, head 2 stays fixed and pulls head 1 forward. Alternatively, in the inchworm model (2) only the forward head catalyzes ATP and always leads while the other head follows (see figure below).

In both of these mechanisms, the motor needs two heads to be able to stay on the track as it moves and its step size depends on the length of the legs. However, myosin VI with short legs (8 nm) was observed to take the same long steps (30 nm) as myosin V. Moreover, a single-headed processive motor has suggested that two heads are not necessary for processive motion. These observations lead to another mechanism: biased diffusion of the motor along the actin/microtubule lattice (3). The bias is provided by the initial push of the power stroke, and the motor most likely attaches to the next binding site in the forward direction. Understanding motor protein movement is a fundamental step in understanding how cargo transport works within a cell, but despite intensive research, the mechanism underlying movement remained highly controversial.

The most direct way to distinguish among these models is to measure how much each head moves when the motor walks. The hand-overhand model predicts that a head alternately moves twice the stalk displacement and stays stationary in the next step while the other head takes a step (see figure, left panel). In contrast, the inchworm model predicts that both of the heads move forward the same distance as the stalk (see figure, right panel). The diffusion model states that heads randomly bind to the track. Current nanometer-precision tracking techniques (optical traps and cantilever probes) cannot readily be used to watch the head movement, because they use a large probe (>100 μm) that might hinder the movement of the motor's tiny heads (5 to 10 nm). What is needed is to track a nanometersized probe (such as organic dyes) attached to a motor head with single-nanometer precision.

The position of a diffraction-limited spot can be localized very precisely by determining the center of its emission pattern. However, organic dyes are not very bright and the signal disappears quickly by permanent photobleaching. This limited previous singlemolecule tracking experiments to a precision of around 30 nm (4). I have extended the photostability and brightness of single organic dyes 20 times by effectively deoxygenating the assay solution and using reducing agents, and I have achieved 1.5-nm localization and collected 1.4 million photons from single organic dyes. The technique, named fluorescence imaging with one-nanometer accuracy (FIONA), has improved spatial resolution in single molecule fluorescence by ∼20-fold.

Using FIONA, I tracked the movement of the motor proteins myosin V, kinesin, and myosin VI, which were labeled with a single dye in the head region as follows.

Myosin V. Bifunctional rhodamine (Br)-labeled calmodulins were exchanged into the myosin V lever arm, where the calmodulin can potentially exchange at any of six calmodulin-binding sites (IQ domains). The inchworm model predicts a uniform step size of 37 nm regardless of the position of the labeled calmodulin. The hand-over-hand model predicts alternating short and long steps, depending on the in-plane distance of the dye from the midpoint of the myosin. The trajectory of moving spots created three classes of steps. I observed 74-or 0-nm displacements for dye on the first IQ domain, alternating 52-and 23-nm steps for dye on the fifth IQ domain and alternating 42-and 33-nm steps for dye on the sixth IQ domain (5) (see figure below, left).

Myosin V: Walking or inchworming?

Predicted movement for the heads and a dye molecule label (green dot) on the lever arm in the hand-over-hand model (left) and the inchworm model (right). The FIONA assay has revealed that myosin V, along with kinesin and myosin VI, walks hand-over-hand.


Kinesin. A human kinesin was specifically labeled on the head region with a single Cy3 molecule. As the stalk took 8-nm steps, the head was observed to take alternating 16-nm and 0-nm steps (6).

Myosin VI. Myosin VI was labeled with a single Cy3 molecule on a calmodulin-binding site. Again, the labeled head alternately moved twice as far as the stalk moved and stopped as the other head moved (7). Unexpectedly, Cy3-calmodulin showed significant flexibility when it had ATP bound, whereas it was immobile in the nucleotide-free and ADP states. This implies that in some part of myosin VI's ATPase cycle, the lever arm uncouples from the motor, which could arise from elongation of the lever arm. Lever arm elongation may provide the long step (30 nm) of myosin VI with a short lever arm (8 nm).

Thus we have established a new, single-molecule fluorescence technique, FIONA, which is able to resolve steps of a few nanometers taken by molecular motors. FIONA assays on myosin V, myosin VI, and kinesin have revealed that these motors move by walking hand over hand, not by “sliding” like an inchworm, nor by “diffusing” along the cytoskeleton. FIONA is also a broadly applicable technique in other fields of molecular biology, such as DNA sequencing and particle tracking in vivo.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.

Navigate This Article