How kinesin actually moves

Recently, Carl Zimmer made a criticism of the computer animations of molecular events (it's the same criticism I made 8 years ago): they're beautiful and they're informative, but they leave out the critical aspect of stochastic behavior that is important in understanding the biochemistry. He's talking specifically about kinesin, a transport protein which the animators are particularly fond of illustrating.

Every now and then, a tiny molecule loaded with fuel binds to one of the kinesin "feet." It delivers a jolt of energy, causing that foot to leap off the molecular cable and flail wildly, pulling hard on the foot that's still anchored. Eventually, the gyrating foot stumbles into contact again with the cable, locking on once more -- and advancing the vesicle a tiny step forward. This updated movie offers a better way to picture our most intricate inner workings.... In the 2006 version, we can't help seeing intention in the smooth movements of the molecules; it's as if they're trying to get from one place to another. In reality, however, the parts of our cells don't operate with the precise movements of the springs and gears of a clock. They flail blindly in the crowd.

The illusion of directed, purposeful movement is a simplifying shortcut: as Zimmer describes, there actually is a lot of noise in the system, it's just that the thermodynamics of the interactions promote a directionality to the motion. This is Chemistry 101. I figured that everyone with an undergraduate level of understanding of molecules would be able to grasp this.

I did not take into account willful ignorance, however. Jonathan Wells is angry that anyone dared to question the perfect "stately grace" of molecular machines, and accuses proponents of stochastic motion of Flailing Blindly: The Pseudoscience of Josh Rosenau and Carl Zimmer. He has a Ph.D. in biology, and he doesn't understand what I just said was Chem 101? For shame.

But that's not what the biological evidence shows. In fact, kinesin moves quickly, with precise movements, to get from one place to another. A kinesin molecule takes one 8-nanometer "step" along a microtubule for every high-energy ATP molecule it uses, and it uses about 80 ATPs per second. On the scale of a living cell, this movement is very fast. To visualize it on a macroscopic scale, imagine a microtubule as a one-lane road and the kinesin molecule as an automobile. The kinesin would be traveling over 200 miles per hour!

The speed of the reaction doesn't say anything about the specifics of the molecular movement…and it's especially not convincing when your trick is to multiply the actual speed in the cell by approximately 1012 to scale it up to the size of a car. The flow rate of the Mississippi river is about 1.5 miles per hour here in Minnesota -- if you multiply that by 1012, oh my god, the water is moving at about 2000 times the speed of light!

But let's set aside the stupid inflation for a minute. Wells cites a couple of papers to back up his claim of the rate of ATP consumption. It's true. But it doesn't show that the movement is steady and machine-like and precise at all. He must be trusting us to not bother even reading the paper.

Here's the deal: we can actually watch single molecules of kinesin behaving. The typical trick is to use a fluorescent bead, attach that to kinesin, and then record the glowing bead's movement as it is moving along with optical-trapping interferometry. That's the problem with Wells' accusation: we actually see the behavior, and it's not linear, smooth, and graceful.

This is the data that the paper used to measure the quantum, jerky behavior of kinesin. Just look at the top graph: that's a record of the bead's movement over time. You should be able to see that the line holds steady at one distance for variable lengths of time, and then jerks upward. The "jerks" are distances of about 8nm, and the other graphs are power spectra to show that there is a peak periodicity of 8nm. It shows the opposite of what Wells claims; there are long pauses and sudden shifts in the directly observed track of kinesin movement. The 8nm emerges because when one "foot" of kinesin releases and wobbles forward to connect to tubulin, it has an 8nm step.

Horizontal grid lines (dotted lines) are spaced 8 nm apart. Data were median-filtered with a window width of 60 ms. b, Normalized histogram of pairwise distances between all pairs of data points in this record, showing a clear 8-nm periodicity. c, Normalized power spectrum of the data in b, displaying a single prominent peak at the reciprocal of 8 nm (arrow). d, Variance in position, averaged over 28 runs at 2 microM ATP (dots), and line fit over the interval 3.5 ms to 1.1 s. The y-intercept of the fit is determined by equipartition,  left fencex2right fence = kT/alpha, where alpha is the combined stiffness of the optical trap and bead–microtubule linkage. The rapid rise in variance at short times reflects the brownian correlation time for bead position. Horizontal grid lines (dotted lines) are spaced 8 nm apart. Data were median-filtered with a window width of 60 ms. b, Normalized histogram of pairwise distances between all pairs of data points in this record, showing a clear 8-nm periodicity. c, Normalized power spectrum of the data in b, displaying a single prominent peak at the reciprocal of 8 nm (arrow). d, Variance in position, averaged over 28 runs at 2 microM ATP (dots), and line fit over the interval 3.5 ms to 1.1 s. The y-intercept of the fit is determined by equipartition, left fencex2right fence = kT/alpha, where alpha is the combined stiffness of the optical trap and bead–microtubule linkage. The rapid rise in variance at short times reflects the brownian correlation time for bead position.

You can find lots of papers with direct observations of kinesin movement. Here's data from another paper that essentially shows that the two 'feet' of kinesin alternate in their movement, because they made recombinant, asymmetric kinesin with slightly different step distances. Again, note the long dwell times punctuated with surges of movement.

Representative high-resolution stepping records of position against time, showing the single-molecule behavior of kinesin motors under constant 4 pN rearward loads. (A) Limping motion of the recombinant kinesin construct, DmK401. The dwell intervals between successive 8-nm steps alternate between slow and fast phases, causing steps to appear in pairs, as indicated by the ligatures. (B) Nonlimping motion of native squid kinesin, LpK. No alternation of steps is apparent; vertical lines mark the stepping transitions. Slow and fast phase assignments, as described in the text, are indicated in color on the uppermost trace of each panel (blue and red, respectively), and the corresponding dwell intervals are numbered. All traces were median filtered with a 2.5-ms window. Representative high-resolution stepping records of position against time, showing the single-molecule behavior of kinesin motors under constant 4 pN rearward loads. (A) Limping motion of the recombinant kinesin construct, DmK401. The dwell intervals between successive 8-nm steps alternate between slow and fast phases, causing steps to appear in pairs, as indicated by the ligatures. (B) Nonlimping motion of native squid kinesin, LpK. No alternation of steps is apparent; vertical lines mark the stepping transitions. Slow and fast phase assignments, as described in the text, are indicated in color on the uppermost trace of each panel (blue and red, respectively), and the corresponding dwell intervals are numbered. All traces were median filtered with a 2.5-ms window.

The contrast is between "stately grace" and "jiggle and jump". The evidence shows that it is the latter. Yet, somehow, Wells closes his weird series of non sequiturs with this question, as if he expects everyone to give him the answer he wants:

So, who are the pseudoscientists?

The answer is obvious. Wells and his cronies at the DI.


Asbury CL1, Fehr AN, Block SM (2003) Kinesin moves by an asymmetric hand-over-hand mechanism. Science 302(5653):2130-4

Schnitzer MJ1, Block SM (1997) Kinesin hydrolyses one ATP per 8-nm step. Nature388(6640):386-90.

More like this

One of the traditional ways to explain a scientific subject is the historical approach: start at the beginning of the endeavor and explain why people asked the questions they did, how they answered them, and how each answer blossomed into new potential. It's a popular way of teaching science, too,…
Many of you have already seen the gorgeous video below: it's a spectacularly beautiful animation of the activity in a cell. I like it, and it's a useful illustration, but … there's something fundamental that it gets completely wrong. So today I'm not going to praise it, I'm going to criticize it…
I was wrong — it's not the Harvard multimedia video. It's an independently generated copy. I grabbed a few images from the DVD I got at my truncated visit to the Expelled screening, and here, for instance, is the segment that shows that striking kinesin motor protein towing a vesicle down a…
As we are so often reminded by proponents of Intelligent Design creationism, we contain molecular "machines" and "motors". They don't really explain how these motors came to be other than to foist the problem off on some invisible unspecified Designer, which is a poor way to do science—it's more…

But that’s not what the biological evidence shows. In fact, kinesin moves quickly, with precise movements, to get from one place to another.

This mental picture - that cellular chemical reactions are "graceful", purposeful, and like well-tuned machinery - is central to the lay-understanding of Intelligent Design. It just intuitively makes sense - how could something so graceful happen by chance, when we see disorder in every day life?

For me, one of my big "aha!"s in believing in evolution came when I realized that the same entropic processes that we see on the macroscale also apply on the microscale, and that cellular movements aren't graceful at all, but chaotic. And they're actually *more* chaotic - we don't have macroscale machinery that works by randomly being jumbled around. Such a thing can't actually happen can't happen at the macroscale - which tells us that macroscale machinery is actually quite different from cellular machinery, in how it operates.

Most IDists take the macroscale analogy of "machinery" down to the microscale, and imagine the same kind of precision, and wonder "how is this supposed to happen by chance?".

What they *should* be doing is taking the chaos that they see in a tornado, scaling it down to chemical reactions, and then learning some science to understand how thermodynamics creates something functional out of the chaos. But it's still chaos, it's not pretty, or fine-tuned, or graceful. The chemical processes in our cells work *by* random jumbling, not in spite of.

I think Wells is demonstrating that he has an overly-idealized picture of how cellular machinery works. That idealization is a fundamental part of why ID persists. Without an understanding of Brownian motion and stochastic process, evolution, heck, even chemistry, doesn't make any sense.

By Windchaser (not verified) on 22 Apr 2014 #permalink

^ Exactly.

By David Marjanović (not verified) on 23 Apr 2014 #permalink

As we are aware the cytoskeleton is composed of three parts : the actin filaments, microtubules and intermediate filaments. More importantly, microtubules forms the 'tracks' for transport or movement of vesicles within a cell. These tracks need a motorproteien in order to connect the vesicle that needs to be transported to the tracks - thus kinesin and dynein (two different types of motorprotein) are needed. The kinesin motorproteins carries their cargo vesicles from the centre of the cell to the peripheral lipid layer surrounding the cell.
This random movement of kinesin molecules agree with how microtubules are spontaneously formed throughout the cell, thus the movement of kinesin being random is in agreement with how the cell forms microtubules randomly and swiftly to nourish certain parts of a cell. In order to understand how randomly the kinesin moves it is important to know that the chemistry of a cell is extremely chaotic and unorganized - in accordance with different cell types, cell needs, surrounding environmental conditions as well as cell function.
The speed at which this kinesin motorprotein moves is also understandable, seeing that cells need to swiftly expell any harmful or toxic substance created or which has penetrated the cell's peripheral wall. The cell production of hormones, ATP or any other product which has to be transported to different regions of the body may also need the kinesin extremely fast, thus it is very beneficial for the cell to move these vesicles at high speed out of the cell to the peripheral wall.

The movement of kinesin is thus extremely well adapted to its specific function, even if it may be random or at extremely high speeds.

Thank you,
Elri van den Berg - u14008425

By Elri van den B… (not verified) on 23 Apr 2014 #permalink

I was astonished at the weirdness of Wells's objections until I realized what the site was. In fact, there are even higher-resolution images of the movement of myosin (a motor protein which crawls along actin filaments) and of course you see the expected randomness in the movement of the free leg (https://www.youtube.com/watch?v=xysUmcsN5DM - this is, I believe, Toshio Ando's work with high-speed atomic force microscopy).

I'm sure you've answered this elsewhere, but I wonder if anyone can point me to an answer: what mechanism makes the kinesin/myosin always step in the same direction rather than random walk?

By Philip Hand (not verified) on 23 Apr 2014 #permalink

Binding ATP makes it bend in a certain direction.

Electrostatics and Brownian motion all the way down.

By David Marjanović (not verified) on 24 Apr 2014 #permalink

Newer movie beautiful - but not sure it helps conceptually even though more accurate. We need visualizations that capture the 'how' of the stochasticity producing the working system - hard

By Debora Marks (not verified) on 24 Apr 2014 #permalink

I think that the movement of Kinesin is amazing and that it is one of the wonderful things that happens in our body everyday. I agree with Dr. Wells that Kinesin moves in precise movments. Although the experimental data shows that Kinesin moves in "jerks" it moves in precise 8nm steps. Kinesin's movements my be random but it is precise and very fast and that is the beauty of nature.

By Anton Venter (… (not verified) on 26 Apr 2014 #permalink

Although the experimental data shows that Kinesin moves in “jerks” it moves in precise 8nm steps.

...That's because of the length of its "legs" and the angles which they can assume. It's not some kind of miracle; indeed, anything else would be a miracle.

By David Marjanović (not verified) on 26 Apr 2014 #permalink

The movement of Kinesin has fascinated me ever since I learnt about it. Even though the experimental data indicates that Kinesin moves in so called "jerky" movements, the actual size of the steps that Kinesin takes are precise. This is another marvellous example of how the human body works in a perfectly structured manner.

By Nikiyta Naude … (not verified) on 26 Apr 2014 #permalink

You're not listening, eh?

By David Marjanović (not verified) on 27 Apr 2014 #permalink

On seeing this blog I was instantly intrigued as when I learnt about the movement of these motor proteins, even on the basic level that is taught in first year, I was amazed by the precision and direction that these motor proteins achieve when we were shown a beautiful animation in a video called " Inner Life of a Cell" of a Kinesin motor protein pulling a vesicle along -3.42 minutes into the video (http://www.youtube.com/watch?v=B_zD3NxSsD8). From my understanding most metabolic process and processes that take place in calls can be described as random as although they are truly amazing and have very specific functions how they achieve what they are interned to is very random and as described in this blog as chaos. This is why the precise movement of the Kinesin protein is somewhat unbelievable and therefore taking the seemingly precise 8nm "step" length and the random "foot" movement as described by this blog , I conclude that the Kinesin protein does take precise 8nm "steps" but how its "foot" gets to the next 8nm position- the movement of the detached foot, is as described in the blog as flailing blindly until it reattaches to the microtubule. Nevertheless this is still one of the amazing processes that we take for granted everyday and sadly that most people do not even know about. Things like this really make you appreciate life whether you see it as random or precise it is still truly amazing.

By Emma Cook (u13… (not verified) on 30 Apr 2014 #permalink

whether you see it as random or precise

This is not something you can choose. It's an observed fact that it's random. Amazing, but random.

The animation you link to is just that: an animation. It's not a recording of how things really happen. It fails to show Brownian motion.

The animation shows goal-directed movements (for example when it depicts the assembly of actin filaments and microtubules, or when the ribosome homes in on the translocator) as if the molecules had eyes and a brain and propulsion systems so they could go precisely where they needed to. Of course that's not how it works! Reality is much, much more jittery.

The animation shows each step kinesin takes as occurring in the same length of time. This very post shows that that's not how it works: the time each step takes is random. Each foot, once it has dissociated from the microtubule, flails around by Brownian motion till it happens to hit the microtubule again and stick to it; it's not surprising that the time this takes varies.

Unsurprisingly, the animation is also simplified in other ways. At the smallest scale, when it shows you the individual lipid of the cell membrane instead of just a liquid, you should see the individual water molecules as Mickey Mouse heads that bounce around like crazy and crash into everything. At one of the large scales, it shows a mitochondrion worming around as if it had muscles; of course it doesn't, it moves only by Brownian motion.

Life doesn't happen at 0 kelvins.

By David Marjanović (not verified) on 30 Apr 2014 #permalink

Wow this is extremely fascinating .
To analyze the velocity - force data and gain insight into the intermediate motion when kinesin molecules takes an 8 nanometer step, discrete state stochast models with a dimensional energy landscape needs to be applied. With all this hard work that needs to be done i would personally give Wells some credit and agree that they are the pseudoscientist even though i find it to be too amazing how a kinesin molecule takes one 8- nanometer step along a microtubule and still the movement of this molecule is specific.
How can it move so precisely and steady as if it is moving in a straight line while the supporting data shows clearly that it moves asymetrycal.
With these questions stuck in my mind and unfilled gaps of information if find here i still think that this is amazing and interesting !!

By Senzekile Maba… (not verified) on 30 Apr 2014 #permalink

Hello, PZ Meyers

This was a truly enlightening read, and one that I believe highlights how dynamic the contemporary understanding of biology at the molecular level truly is. Challenging even the most innate and convenient of modern theories and hypotheses is one of the mechanisms allowing science to develop.

Regarding the kinesin motor, specifically: the second law of thermodynamics (any transfer or conversion of energy invariably results in some of the energy being lost as heat, and thus being 'wasted') would imply that erratic movements of the kinesin motor would result in a decrease in efficiency, as more energy would be lost as external radiation. The model of the kinesin motor gracefully completing its movements in precise steps would make sense in such a scenario... Yet, as you have pointed out, whilst adaptation may move an organism or system towards idealism or perfection, it is still an imperfect system nonetheless, and, according to our current understanding, will remain as such.

I suppose this raises the question: whilst we naturally are inclined to eliminate as many assumptions as possible when forming theories and select the 'simplest' hypothesis as the favoured one, how much accuracy do we risk losing in choosing only that which makes the most sense?

Do you think this is an example of a model that requires adjustment precisely because it oversimplifies a concept?

Thank you,
Ryan R.
Undergraduate Student

By Ryan R. (u14086078) (not verified) on 03 May 2014 #permalink

Hello, PZ Meyers

His name is Myers, and he practically never seems to take a look at the comments in this version of Pharyngula – the full version is at http://freethoughtblogs.com/pharyngula/ .

the second law of thermodynamics [...] would imply that erratic movements of the kinesin motor would result in a decrease in efficiency, as more energy would be lost as external radiation.

I'm not sure. This is Brownian motion (both of the protein itself and of the water molecules around it); it is inevitable at any non-negligible temperature. Each step is powered by the hydrolysis of an ATP molecule; a smaller amount of energy is not available in an Organism As We Know It.

how much accuracy do we risk losing in choosing only that which makes the most sense?

This is a testable hypothesis that must be tested by observation. :-) That's exactly what, for this case, Asbury et al. (2003) have done.

By David Marjanović (not verified) on 03 May 2014 #permalink

I’m sure is question must have been asked, although I have been wondering if anyone can point me to an answer: what mechanism makes the kinesin/myosin always step in the same direction rather than random walk?

The negative charge from the extra phosphate, which it takes from ATP, deforms it in a way that makes it step forward. The random part is the next attachment to the microtubule.

By David Marjanović (not verified) on 05 May 2014 #permalink