Molecules: A Very Short Introduction - Ball Philip 2001

Good little movers: molecular motors

After-dinner speeches are not normally notable for launching revolutions. But Richard Feynman, who was engaged to address the West Coast section of the American Physical Society in 1959, was not a normal physicist. One of the most creative scientific minds of the post-war twentieth century, he is most vividly remembered by the world at large as a bongo player, a practical joker, a safe-cracker, the trickster figure of modern science.

Feynman’s talk in 1959 was high-spirited but ultimately serious in its intent. He called it ’There’s Plenty of Room at the Bottom’, and it was about engineering on scales too tiny to see. ’What I want to talk about’, he said, ’is the problem of manipulating and controlling things on a small scale’. By ’small’, said Feynman, he did not mean ’electric motors that are the size of the nail on your small finger’. He meant small as in atoms.

’Imagine’, he went on, ’that we could arrange atoms one by one, just as we want them’. This, he saw, is essentially what the chemist tries to do:

The chemist does a mysterious thing when he wants to make a molecule. He sees that he has got that ring, so he mixes this and that, and he shakes it, and he fiddles around. And, at the end of a difficult process, he usually does succeed in synthesizing what he wants.

You can see that the physicist’s view of what chemists do is hardly more sophisticated than a lay person’s. But Feynman’s description is really not so different from the one Primo Levi gives when explaining how chemists build molecules as engineers build bridges (see page 24). However, the chemist is traditionally accustomed to regarding his molecule as a substance, something to crystallize and put in a bottle. The physicist, on the other hand, sees it as a construct, like an engine component.

Feynman was essentially wondering whether physicists might figure out how to do what chemists do, but wearing an engineer’s hat. Can we build a molecule by pushing atoms into place, one by one? In 1959 such a thing was unthinkable to anyone but a conjuror of the imagination like Feynman.

Yet he was not simply speculating idly. Even at that time, it was clear that technology was getting smaller. The invention of the transistor in the 1940s had shrunk the scale of electronics. Bulky boxes filled with vacuum tubes had been replaced by compact devices containing ’solid-state’ circuits made from silicon transistors. The portable transistor radio was on every American beach. Engineers were becoming increasingly skilled at making tiny machine components — much more so, in fact, than Feynman realized. Hoping to provide some small impetus for driving miniaturization technology forward, he offered two prizes of a thousand dollars, to be funded by himself: one for making an electric motor measuring no more than 1/64th of an inch on any side, the other for writing the information from a page of a book in an area scaled down by a factor of 1/25,000. Feynman presumably anticipated that his money would be safe for some years to come — he did not imagine that someone (an engineer named William McLellan) would meet his first challenge within a few months.

Today we can go further. Tiny cogs and motors a tenth of a millimetre across have been carved out of silicon wafers using acid etching or electron beams (Fig. 25). But carving out parts from slabs of material is all very well until you reach a scale of around a tenth of a micrometre: current methods for making integrated circuits in silicon can just about make wires this thin. Beyond that these methods cannot go — it becomes like trying to split a human hair with a bread knife.

Researchers are starting to ask whether this ’top-down’ approach still makes sense at such scales. Components this small are closer in size to molecules (medium-sized molecules are a hundred times smaller) than to silicon wafers you can hold and see in your hand. Should we then start making things from the bottom up — from single molecules?

Primo Levi confessed in The Monkey’s Wrench that chemists fantasize about a tool kit for molecular-scale construction:

we don’t have those tweezers we often dream of at night, the way a thirsty man dreams of springs, that would allow us to pick up a segment, hold it firm and straight, and paste it in the right direction on the segment that has already been assembled. If we had those tweezers (and it’s possible that, one day, we will), we would have managed to create some lovely things that so far only the Almighty has made, for example, to assemble — perhaps not a frog or a dragonfly — but at least a microbe or the spore of a mold.

Feynman too found inspiration in the molecular devices and artefacts of biology: ’in which chemical forces are used in a repetitious fashion to produce all kinds of weird effects (one of which is the author).’ He realized that there are already molecular machines in biology. In 1959 this, had it reached the ears of biologists, would doubtless have been dismissed as the attempt of some foolish physicist to impose his own perspective on a field he clearly knew nothing about. But today biologists are quite happy to talk about proteins as molecular machines.

25. A micromotor carved from a silicon wafer.

This chapter looks at some of the most remarkable of these: protein molecules that create movement. They are molecular motors, often called motor proteins. The mini-motor that won Feynman’s prize is a clumsy gargantuan in comparison, as a lumbering Diplodocus is to the nimble flea. The biological importance of motor proteins is immeasurable. Without them, we could not move a muscle; no birds would cross the sky, no fish ply the seas. Even bacteria would be immobile. But worse: cells could not divide, so there would be no reproduction. Without molecules to drive movement, there is no life.

Yet for the mechanic of the molecular world, motor proteins say something else. They show that molecular-scale engineering is possible: that we can scale down ideas familiar from the everyday world to the realm of molecules. Motor proteins are not unique in this respect, but they make the point with rare explicitness. I will describe how we might achieve similar goals from scratch, by making our own custom-built molecular motors. This leads us into the arena towards which Richard Feynman’s talk was the first clear signpost: the science of nanotechnology, which is technology on the scale of nanometres — distances one can measure in molecules.

Front crawl

The shape of a molecule is never fixed: it is always vibrating and waggling its loose, floppy parts. Mechanical motion is ubiquitous in the molecular world.

Yet, in general, either molecular motions are random, like the meandering wriggle of a polymer chain floating in solution, or they average to zero, like the back-and-forth vibrations of a chemical bond. What we need from a genuine motor, in contrast, is motion with a directional bias — what one might call purposeful motion.

Any motor consumes fuel. You could regard this as the inevitable price of orderly motion, a toll imposed by the Second Law of Thermodynamics. Random molecular motion, on the other hand, can be had ’for free’ — it is the incoherent molecular jiggle that is heat.

Our bodies conduct many types of directional transport. For example, the motion of cilia — hairlike appendages that line the air passages of our lungs and windpipe — moves a layer of mucus from the lung lining up to our throats, where it accumulates as phlegm. This mucus captures dirt, and so its export keeps the lungs clean. To move the mucus up the windpipe, the cilia cannot just thrash around but have to execute a coordinated sequence of movements, like the arms of a swimmer. They make a whiplike ’power stroke’ followed by a slow, crawling ’recovery stroke’. Some single-celled organisms called protists do in fact use cilia on their cell surface to swim through water.

The molecular engine that drives these motions is a protein called dynein. Each cilium contains microtubules (see page 63) arranged around the circumference of a tube called the axoneme. The tubules are joined together in pairs called doublets, like the barrel of a twin-barrelled shotgun. There are nine paired microtubules per axoneme, and they are interconnected by dynein molecules, protruding at regular intervals like the legs of a millipede (Fig. 26).

To move the cilium, the microtubules walk over one another. Each dynein molecule has a ’leg’ that bends by undergoing a reaction that consumes ATP. Dynein is basically an enzyme that breaks apart ATP and changes shape as a result. Calcium ions are also needed to trigger this reaction. The motion is controlled by nerve signals that trigger the injection of calcium into the cilium.

26. The bending of cilia is driven by dynein molecular motors.

Since the dynein molecules all point in the same direction, they pull one microtubule doublet over another when they bend. If the molecules were then simply to straighten again, the microtubules would return to their original positions. In order to generate forward motion, each dynein molecule detaches itself from the second microtubule before straightening up, and then reattaches for the next ’power stroke’. Only when its ’foot’ is attached to another microtubule can dynein break down ATP and switch to the bent state.

The crawling of doublets one over the other is thus like a ratchet: the cycle of attachment, bending, detachment, and straightening of dynein generates motion in a single direction. But, because the ends of the microtubules are anchored at the base of the axoneme, this sliding of one over another causes bending of the cilium. With proper coordination of the sliding motions, the cilium will bend first this way and then that. The coordination seems to come from a pair of microtubules running down the centre of the axoneme, although exactly how they achieve this is not yet understood.

Dynein plays a broader general role in the world of the cell: it is one of the engines that shuttle objects around. Our cells are laced with an internal rail network of microtubules. From time to time the cell needs to rearrange its compartments, the membrane-branded structures called organelles. Attached to a membrane wall, dynein can pull an organelle along the tracks.

These journeys are one-directional. The ends of microtubules are not equivalent: only from one of them, called the plus end, can tubulin molecules (see page 64) be added or removed. Dynein always moves towards the other extremity — the minus end — of a microtubule, which lies towards the cell’s centre. When a cell divides in two, dynein pulls the duplicated sets of chromosomes along the microtubules of the mitotic spindle (see page 65), carrying them towards the centres of the respective nascent daughter cells.

For transport along microtubules in the other (plus) direction, a different motor protein called kinesin is used. Kinesin is perhaps the most anthropomorphic of molecules that induce motion, since it has two ’legs’ and executes a waddling ’walk’ in comparison with dynein’s one-legged inchworm crawl. Kinesin too is powered by a reaction that consumes ATP and alters the protein’s shape.

Kinesin is the cell’s postman, delivering parcels from one organelle to another. For example, proteins must be sent from their point of manufacture (the endoplasmic reticulum) to the parts of the cell where they are needed. They are packaged inside little membrane spheres called transport vesicles, and kinesin carries a transport vesicle along the microtubule network to the right address.

Muscle power

Our own walking is enabled by muscle contractions and extensions. We are sprung and countersprung with so-called skeletal muscles, which, simply by shortening and relaxing, can control everything from a pianist’s delicate finger movements to the pounding of an athlete’s thighs.

Skeletal muscle is one of nature’s hierarchical molecular materials (see page 53). It is a fibrous composite of bundles within bundles within bundles. Individual muscle cells are extremely elongated and enclose many-stranded cables woven from threads called myofibrils. Within the complex molecular substructure of these strands reside the secrets of muscle contraction.

Seen through the microscope, a myofibril is punctuated by light and dark bands of various widths. These give skeletal muscle a striped appearance at high magnification, which is why it is also known as striated muscle. The sequence of bands repeats periodically along the myofibril strand, and one repeat unit is called a sarcomere. The various bands within a sarcomere are given the kind of anodyne names — A band, H zone, and so forth — that are always telltale indications that no one had the faintest idea, when they were first observed, what they signified.

Noticing in the 1950s that these bands changed width when muscle contracted, Andrew Huxley, Hugh Huxley, and their co-workers proposed the so-called sliding filament theory of muscle action. Their idea was that the myofibril contains toothbrush-like structures facing one another and pushed together so that their bristles interpenetrate. Each sarcomere contains a double set of these pairs of brushes, placed back to back. The dark bands correspond to regions where the bristles interdigitate (creating a high density of molecules), whereas in light bands there is only one set of bristles (Fig. 27). The Huxleys suggested that the myofibril shortens — and so the muscle contracts — by deeper interpenetration of the sarcomere’s bristles.

27. Interdigitating filaments in muscle allow it to contract.

This movement of filaments over one another is driven by the motor protein myosin, a long thin protein in which two helical chains twist around one another. At either end, the chains terminate in a pear-shaped head. Myosin molecules are gathered into bundles called myosin filaments. Each end of a filament bristles with myosin heads, like a head of corn budding from a sheaf.

Penetrating between the myosin bundles are filaments of a protein called actin. This protein is in fact globular in shape, but the globules link together to form a chain like beads on a necklace. In an actin filament, two chains of actin ’beads’ wrap around each other in yet another double helix. The necklace is further adorned by strands of the protein tropomyosin, which wind their way along the actin filament. And at regular intervals sits a globular protein called troponin (see Fig. 27).

Muscle contracts when myosin heads attach themselves to the actin filaments and pull themselves along. The principle is the same as that by which dynein and kinesin move along microtubules: motion is generated by a change in shape of the attached motor protein, driven by the breaking-down of ATP to ADP. The myosin head swings on a hinge connecting it to the rest of the molecule. It kinks, detaches itself from actin, unkinks, and reattaches, and thereby ratchets along the actin filament in a series of power strokes (Fig. 28).

The whole process is under voluntary control, set in motion when a nerve impulse from the brain tells the muscle to tighten or relax. The tropomyosin and troponin proteins on the actin filament provide the switch. Muscles are wired to the brain by nerve cells called motor neurons, which are like wires biochemically ’soldered’ to the outside of a muscle fibre. An electrical signal arriving at the end of the motor neuron triggers the release of calcium ions from a web of tubes — the sarcoplasmic reticulum — that extends through the spaces between myofibrils inside the muscle fibre. These calcium ions are captured by troponin molecules on the actin filament, prompting the proteins to change shape. This in turn pulls on the tropomyosin strands, which twists the double-helical actin necklace and rotates the actin beads. It is this rotation that exposes the sites on actin to which myosin binds. The entire structure thus contains an elegant molecular transmission mechanism for switching contraction on and off.

28. Muscle contraction is caused by the movement of myosin molecular motors along filaments of the protein actin.

Molecular tweezers

Once upon a time molecular scientists had to deduce all they knew about molecules from measurements made on many billions of them simultaneously. This can be a risky business, since we cannot always be sure how such measurements are related to the properties of individual molecules, just as the noise that emanates from a football stadium or theatre hall reveals nothing of the individual conversations people are having. But advances in experimental techniques that enable studies of single molecules — what they look like, how they interact, how they move — have over the past two decades opened up an entirely new realm of molecular studies. We are starting to get to know molecules in person.

One of the critical innovations is the invention of tweezers for molecular manipulation — the very tool that Primo Levi desired. The most remarkable thing about these tweezers is not that they are so fine but that they are literally intangible — they are made of light. They are called optical tweezers, and they trap objects in a very intense light beam. They allow researchers to address the kinds of questions one might ask about everyday mechanical motors: how efficient are they, how much load can they bear, how fast do they move?

Interaction between light and the electrons in molecules can create a force — a kind of ’light pressure’ — on an object. If the object is small enough and the light intense enough, the object can be moved by this force. In optical tweezers, the intersection of two or more laser beams sets up a spot of extremely bright light. A small object within this bright pool experiences a light pressure from all sides that prevents it from moving in any direction. It is caught in an optical trap between the tweezers of the laser beams. If the beams are moved, the object is pulled along with them.

The force generated by a single motor protein can be measured by tethering either the motor or the object it moves along (an actin filament, say) to a microscopic plastic bead clamped between optical tweezers. Motion generated by the motor tugs the bead away from the centre of the trap by an amount proportional to the force generated.

Using a bead as a handle, optical tweezers can be used to do extraordinary things with molecules. Kazuhiko Kinosita at Keio University in Japan and his co-workers attached beads to each end of an actin filament and then pulled one end hither and thither until it was threaded through a loop, creating a molecular knot (Fig. 29). They tightened the knot until it broke. Because the actin filament is somewhat stiff, like a sapling branch, it is weakened when sharply curved, and the force required to break the knotted filament was far lower than that needed to pull apart an unknotted filament.

Optical tweezers are not the only tool for handling molecules one at a time. Devices called scanning probe microscopes, devised in the 1980s (and used to take the images shown in Fig. 5), have proved immensely valuable not just for observing but for manipulating the molecular world. One of these instruments, called the atomic force microscope (AFM), allows researchers to probe the mechanical properties of molecules — how stiff or stretchy they are, for instance. A molecule can literally be grasped at one end by the AFM and pulled like a piece of elastic.

29. Optical tweezers were used to tie this knot in a strand of actin. The microscopic beads attached to each end acted as ’handles’. The actin is made visible under a microscope by shining light on it to make it fluoresce.

Motors by design

One of the most prominent prophets of nanotechnology is K. Eric Drexler, an independent scientist who heads the Foresight Institute in California. Drexler’s vision, which is built around the idea of molecular-scale robotic assemblers that can put together any molecular machine (including themselves) atom by atom, has been influential on the public perception of nanotechnology’s promise (and dangers), but rather less so among scientists. Some scientists worry that Drexler’s idea of atom assemblers fails to take account of heat that must inevitably be released when atoms are combined. Moreover, the shapes of molecules are many and varied, but not arbitrary: there is no guarantee that a particular molecular-scale blueprint for a nanotechnological component will correspond to a stable or realizable arrangement of atoms.

Drexler first outlined his ideas in his 1986 book Engines of Creation, in which the protagonists (and sometimes antagonists) were nanotechnological robots. But in terms of what is already technologically feasible, a scratch-built, controllable molecular motor would in itself be an engine of creation, however primitive. With such a device, molecular-scale rods, girders, and other construction parts might be shifted into place ready for welding together.

Whereas motor proteins are powered by ATP, some researchers think that synthetic molecular motors could be light-powered. In 1999 a team of chemists led by Ben Feringa at the University of Groningen in the Netherlands devised a molecular rotary motor, in which a rotor spins in a single direction driven by light. They exploited the process of photoisomerism: the light-induced interconversion of two different forms (isomers) of a molecule, which have the same chemical constitution but different shapes.

They constructed a molecule containing two linked propeller-like units (Fig. 30). Initially the propellers sit on opposite sides of the molecule: the so-called trans isomer. But ultraviolet light converts the molecule to the cis (’sis’) isomer, in which both propellers are on the same side. So as not to bump into one another, the propeller blades twist — one upwards, one downwards. If the molecule is warmed up above 20 °C, the blades switch to the opposite configuration: that which twisted downwards now bends upwards, and vice versa. In this configuration the molecule is slightly more stable. Irradiating it with another dose of ultraviolet light then brings about the reverse switch from the cis to the trans form. But because of the propeller flip that preceded it, the trans form is now subtly different from that at the start: the propellers both bend down rather than up. Heating the molecule to 60 °C restores the original configuration.

The overall result of this four-step process is that one of the propeller blades makes a full revolution with respect to the other, in a predetermined direction. If the molecule is kept above 60 °C and continuously irradiated with ultraviolet light, it will spin smoothly: a light-powered molecular motor.

A different rotary device was made by Ross Kelly and co-workers at Boston College. They constructed a molecule consisting of a three-bladed propeller connected by an axle to a ’brake’ that hindered the rotation of the propeller. Without the brake, the propeller rotates — but at random in either direction. The researchers aimed to use the brake to ratchet the propeller in only one direction, by executing a series of chemical reactions between blade and brake. But they have not yet found a way to pull their prop through more than a third of a full turn.

30. A scratch-built molecular rotary motor powered by light. The top image shows the carbon-atom framework.

Both of these devices are simplistic and neither can yet be harnessed to perform a useful task. But they show how, in principle, molecular motors might be constructed. The sequence of bond making and breaking required to move Kelly’s motor looks cumbersome, but a similar sequence is after all needed to produce linear motion with kinesin and myosin. At the molecular scale, such things can happen quickly enough to give the appearance of smooth motion.

Natural nanotechnology

Synthetic molecular motors have a long way to go before they can compare with natural motor proteins. Does it, then, really make sense to try to build them from scratch, or might one instead adapt motor proteins to nanotechnological ends? Some researchers have isolated motor proteins from the cell and chemically modified them so that they can perform new tasks.

In 1997 Stanislas Leibler at Princeton University and co-workers made devices from the motor protein kinesin that could arrange microtubules into organized patterns. They linked four kinesin molecules together chemically, forming an assembly rather like a creature with four sets of legs. When mixed with microtubules and fed with ATP, these kinesin constructs pulled the tubules one past another until they became organized into star-shaped structures (Fig. 31) very much like those formed in the first stages of cell division (see page 65).

31. Microtubules organized into star-shaped structures by semi-synthetic molecular motors made from modified proteins.

At the University of Washington in Seattle, Viola Vogel and co-workers have used kinesin to propel microtubules over surfaces in a selected direction. They attached kinesin molecules to a surface coated with the polymer polytetrafluoroethylene (PTFE), better known as the non-stick coating Teflon. The PTFE was applied by rubbing a block of it over the surface, whereupon the polymer film acquires striated grooves and ridges in the direction of rubbing. The polymer chains are thought to be aligned with these ridges. Kinesin molecules become attached preferentially on the ridges, which means that they form oriented rows. These rows act as linear tracks along which microtubules can be passed: the kinesin molecules pass the tubules to one another like a bucket brigade. In cells it is the kinesin molecules that are mobile and the tubules that are ’fixed’. But in these experiments the motor proteins are tethered to the surface, so their walking motions propel the microtubules instead.

The most dramatic and exciting amalgamation so far of biomolecular motors with artificial microengineering was described towards the end of 2000 by Carlo Montemagno and co-workers at Cornell University in Ithaca, New York. They commandeered a molecular rotary motor to turn a tiny metal propeller about 150 nanometres wide and nearly ten times as long. The enzyme ATP synthase, we saw earlier, has a head that spins on a membrane-bound spindle as it performs its task of converting ADP to ATP (see page 83). Montemagno and colleagues fixed this head to the top of a microscopic pedestal etched from nickel metal, and then they attached the metal propeller to the spindle. Under the right conditions, ATP synthase can work in reverse, breaking down ATP to ADP and spinning as it does so. The researchers initiated this process by feeding their rotors with ATP, and saw them revolve under the microscope at around five revolutions per second (Fig. 32).

Studies like this raise the exciting prospect of using molecular motors to move molecules around in a controlled way — something that brings a whole new dimension to synthesis at the molecular scale. No longer would chemists have to rely on the random wandering and chance encounter of molecules floating in solution: they could instead guide them precisely where they are meant to go. Because nature has already devised a wondrous array of molecular machines for such purposes, I suspect that molecular nanotechnologists will increasingly make use of the cell’s machinery rather than trying to design devices from first principles. This applies not just to the generation of mechanical motion but to areas such as energy generation, sensors, and information processing. We may then see a fusion of biology with disciplines once regarded as quite different, such as mechanical and electronic engineering. Because the union will bring about results that none of the fields can achieve on its own, we could call it biosynergetic engineering.

32. A microscopic metal propeller attached to the spindle of the rotary motor protein ATP synthase (a) rotates when the motor is driven with ATP. b shows an array of many such constructs. Only those for which the propellers appear here as non-vertical are working as intended: the synthesis is not yet successful in every case.