Walking The Line

Two-legged protein motors are currently marching around your cells delivering cargoes to precise locations, and with impeccable timing. These motors are transporting energy-producing mitochondria to cellular neighborhoods in need of fuel. They are providing the pulling power needed to separate chromosomes during cell division. Motor proteins are even delivering neurotransmitters to the tips of nerve cells, up to 1 meter away from where the neurotransmitters were produced.

"The body of some nerve cells extends from the small of your back all the way to your big toe," according to Steven M. Block, a professor of both biology and applied physics at Stanford University. Motors called kinesins have to walk all the way down the nerve cell to make sure the axon is well stocked with neurotransmitters.

The kinesins "also carry along the reverse motor, so they can walk the vesicle in the reverse direction back to the main cell body to pick up a new load," Block added. "These motors are really amazing."

For example, relying on simple diffusion to get neurotransmitter vesicles to the end of the nerve cells would take years. With kinesins, delivery happens in a matter of minutes. The motors walk up to 100 steps at speeds that can reach 4 µm/second. Like a relay race, the cargo is passed along when another motor is encountered. Sometimes multiple motors team up to carry a single load.

Kinesins are just one of a myriad of protein motors that use chemical energy to do physical work. There's the ribosome, which uses adenosine triphosphate (ATP) to ratchet along messenger RNA to build protein, one amino acid at a time. And viruses use protein motors to stuff their genomic DNA into a protective protein shell. In contrast to these motors, however, so-called processive protein motors are characterized by the ability to use ATP to walk along filament roadways with a load in tow. Such processive motors include members of the kinesin, myosin, and dynein motor families.

Of these, myosins, which move along actin, a polymer roadway that also gives a cell its overall shape, have been studied the longest, mostly because a nonwalking version powers muscle motion. But the stars of myosin processive motion are myosin 5 and its runty cousin myosin 6. Both of these motors take similar-sized steps but walk in opposite directions, each detecting an intrinsic asymmetry of the actin polymer underfoot.

Kinesins and dyneins mostly traffic their payloads in opposite directions, too, but choose a different cellular corridor: namely, thick, hollow, cylindrical filaments called microtubules, which function as the primary transport highways in cells and are essential parts of ciliary and flagellar movement.

Dyneins tend to walk toward the center of the cell, moving payload-filled vesicles recently engulfed at the membrane toward the nucleus. They also help neurons migrate in mammals. Meanwhile, most kinesins stroll outward, delivering vesicles to the membrane for export, moving granules of RNA around, or pulling apart chromosomes during cell division.

Together, kinesin and dynein position organelles strategically throughout the cell, permitting heterogeneous organization as exemplified by neurons. The dendrites, protrusions that receive signals from other neurons, need different protein components than do axons, which send signals out from a long, thick extension. But motors can also facilitate homogeneous dispersal, for example, uniformly distributing pigment in skin cells.

Cargo notwithstanding, a main difference between motors that traffic along microtubules and those that prefer actin filaments is the scale of their footwork. A microtubule's diameter is about three times that of actin, but the motors that walk on a microtubule's wide girth take baby steps, tiptoeing along 8 nm at a time. By comparison, processive myosin motors take long 36-nm strides on slim actin tightropes.

Most marching myosin and kinesin motors are composed of a dimer of two legs, which twist around each other in an α-helical coiled coil. At the top end of this coiled coil are protein domains that link specific cargo to specific motors. At the other end of the motor are two feetlike domains that interact with the filament. These domains are also the catalytic core of these complex enzymes. They each bind and hydrolyze ATP, which in turn leads to the conformational changes that propel the motor forward.

But what coordinates the motion of these two feet in the absence of a brain is a big issue in this field.

"Motor proteins cannot lift up both feet at the same time or they will fall off the filament and diffuse away," said Jon F. Kull, a chemist at Dartmouth College who studies the crystal structures of these motors. "The feet need to coordinate."

"There's something in the step of the left foot that controls the right," Block pointed out. "It's like a dance; somebody's got to lead. But who and how?" Or in structural terms, how does one foot sense the ATP state of the other foot and behave accordingly? Furthermore, how do the motors coordinate so that they walk forward and not simply step back and forth in place, using up ATP without getting anywhere?

Questions like these were elements of the motor minutiae discussed by an eclectic assortment of biophysicists and cell and structural biologists who gathered in late October at the hard-to-resist beachfront conference center in Asilomar, Calif., for the Biophysical Society's biennial Biophysical Discussions meeting.

Although consensus in this field is a rarity-congenial controversy seems like a cultural norm among these scientists and was plain to see in Asilomar-most researchers agree that myosin and kinesin motors move forward in a fashion that resembles the human gait: Every swing of a leg traverses twice the net distance traveled by the motor's center of mass.

Although researchers are still sorting out the nitty-gritty of this walking motion, some mechanistic details are beginning to come to light. "In the past 10 to 15 years, a lot of exciting tools have been developed to study these motors," said Ronald Vale, a Howard Hughes Medical Investigator at the University of California, San Francisco.

These tools include fluorescent probes with nanometer resolution that can be attached to different parts of the motors. There are also optical traps that can determine the strain and other physical properties of motion. And improvements to X-ray crystallography and electron microscopy have been combined to explain how, for example, myosin 5 can make a move along actin when the actin-binding site is 40 Å away from its ATP-binding fuel tank.

It turns out that the crux of myosin 5's motion starts with a glycine residue in a loop that borders the ATP-binding site. When ATP is bound, this glycine moves toward the protein core by 4 Å as it forms a tight hydrogen bond with ATP's γ-phosphate. This small motion is thought to pull on a nearby α-helix, which in turn yields a 60° swing of myosin 5's lever arm. This carefully orchestrated choreography leads to motor propulsion and hinges on a single γ-phosphate, Kull noted. "Motors are effectively a γ-phosphate sensor."

Myosin 5's walking mechanism is relatively intuitive, with a clear link between ATP presence and the physical swing of the lever arm. But its sister protein, myosin 6, whose lever arm is stumpy in comparison, can still take the same lengthy steps. To boot, myosin 6 marches slightly faster, too, said Anne Houdusse, from the Institut Curie, in France. Last year she solved myosin 6's first crystal structure. "This little protein is very curious," Houdusse said. "Every time you get new data about the protein, you also get new puzzles."

"Myosin 6 is really going to reveal how myosins work because we have to explain some really unusual behavior," noted James A. Spudich, a biochemist at Stanford University.

Some researchers believe that myosin 6's motor mechanism will be more like kinesin's. Kinesin shares a core structural topology with myosin; that is, they both boast a conserved central β-sheet and a similar ATP-binding site. "Myosin and kinesin have virtually the same engine," Kull said. "Around this, they have built up their respective functions, just like you can build a car or a boat" around the same V8 engine. In this case, myosin is the boat, at twice the size of kinesin.

Like myosin 5, ATP binding in kinesin pulls a helix toward the core of the protein. But because kinesin lacks myosin 5's long arm-the one whose 60° swing propels the protein forward-ATP binding instead causes a conformational change in the back foot that "cocks" the kinesin dimer and positions the front foot forward on the microtubule. This contortion creates strain that is relieved by release of the back foot, which is triggered by ATP hydrolysis. The way in which kinesin lurches its back foot up off the microtubule and forward is what some researchers call "back- seat driving."

And then there's dynein. This 1,500-kDa monster is about 10-fold larger than kinesin and has no apparent structural similarity to other processive motors, except that it is also a dimer. Unlike myosins or kinesins, each dynein monomer has not one, but four ATP-binding sites. The protein has not yet been crystallized, mostly because it took until this year to coax any organism to produce enough of it for biochemical characterization. Bacterial systems failed, but finally yeast acquiesced, thanks to the coaching of Samara Reck-Peterson, a postdoc in Vale's lab.

Because dynein walks along microtubules like kinesins, it is not surprising that they share an 8-nm step size. "But dyneins also take longer steps, backward steps, and sidesteps," Reck-Peterson said. And though many think that dynein, like myosin and kinesin, walks by alternating its feet, others have their doubts, because one of the motor's legs is 16 nm in diameter, twice the length of the dynein's step size. "Dynein's walk is probably more of a shuffle," she said.

Figuring out how dynein works is one of the next milestones for the motor field, said Ron Milligan, a cell biologist at Scripps Research Institute. "That's the burning question in my mind."

Another outstanding issue in this field is to structurally characterize how motor proteins interact with their filament highways. Electron microscopy has provided several nanometer-resolution pictures of these motors on a filament-enough to begin to see α-helices, but nothing smaller. "Everyone wants to see [atomic-resolution structure] of a motor with a filament," Vale noted.

Chemists are involved in studying the structural biology of these motors and their regulatory mechanisms, while others are designing mechanistically informative synthetic chimeras or modeling motor-protein motion through computational chemistry. A few, like Tarun Kapoor, are developing small-molecule inhibitors.

Kapoor was a postdoc in the lab of Timothy Mitchison at Harvard University when he published the first small-molecule inhibitor of a kinesin that's essential for coordinating the spindle fibers needed for a cell to divide. Just last year, in his own lab at Rockefeller, he went on to characterize how this molecule-dubbed monastrol-works. An allosteric inhibitor, monastrol prevents the conformational changes the kinesin needs to move.

A biotech firm in San Francisco called Cytokinetics is also coming up with kinesin inhibitors and has nine that are currently in Phase II clinical trials against various cancers, including breast and lung cancer. Recent patent applications by AstraZeneca, Merck, Bristol-Myers Squibb, and Chiron are evidence that other drug companies think that these motor proteins are viable drug targets, too.

"Kinesin inhibitors are a very competitive space," said Robert I. Blum, president of Cytokinetics. "We are several years ahead, but there are other companies in Phase I." Merck, for example, completed a Phase I clinical trial last June for a mitotic kinesin inhibitor.

Kinesin inhibitors promise to be more selective than current cancer drugs that target assembly of microtubules (the kinesin roadway) in order to block rapidly dividing cells, Blum argued. Because microtubules are ubiquitous and important for cross-cell trucking of cargo in many kinds of cells, blocking them can have dramatic side effects. But by targeting kinesins involved solely in cell division, it may be possible to avoid such side effects. Cytokinetics is also targeting fungal kinesins, in hopes of blocking infections in immunocompromised individuals.

Indeed, the role of motor proteins in disease is wide-ranging. Since a cell relies on protein motors to compartmentalize its parts, and in some cases to move cellular components quickly, defects in processive motors can lead to hearing problems, left-right asymmetry impairments, and photoreceptor degeneration and kidney disease.

Protein motors even have a connection to Alzheimer's disease. It turns out that kinesins traffic vesicles filled with the amyloid precursor protein, as well as the enzymes that convert it into the amyloid β-peptide, aggregation of which is thought to cause Alzheimer's. Some researchers have suggested that this nefarious processing may occur in the vesicles being actively transported by kinesins.

Processive motors are also hijacked by nasty pathogens that co-opt their roadways to facilitate infection. Because dynein makes a beeline for the nucleus from the lipid membrane, viruses, including HIV and herpes simplex-1, the cause of cold sores, piggyback on this motor protein for quick and strategic delivery to the nucleus, home to the host cell's genetic machinery. Bacteria such as Salmonella also are guilty of hijacking processive motors during infection.

With such a ubiquitous role in cellular trafficking and such a remarkable engine for doing so, there's no limit to the places these motors may carry the research field that studies them.