JUST a few kilometres from the towering Space Needle and the bustle of the
Seattle Center鈥攚here the monorail carries tourists along an elevated
trackway鈥擵iola Vogel is building the world鈥檚 smallest train set. Apart
from an endless supply of ingenuity, this requires a careful choice of raw
materials. And in her lab at the University of Washington, Vogel has discovered
the perfect combination: Teflon and cow brains.
Her trains are made from fragments of microtubules, protein filaments one
thousandth the diameter of a human hair that crisscross the inside of nerve
cells鈥� including those in cow brains. Slice these filaments into minute
segments, drop them onto thin Teflon tracks and the tiny trains race off. 鈥淲e
are learning how to engineer a monorail on a nanoscale,鈥� says Vogel, a physicist
turned bioengineer. 鈥淲e want a molecular shuttle that moves from point A to
point B and which can be loaded and unloaded.鈥�
Her nanotrains are not only fun, they may also be central to the next
industrial revolution, as predicted by techno-visionary Eric Drexler. In his
1986 book Engines of Creation, Drexler describes a world in which
molecular machines take the place of factories. These tiny 鈥渁ssemblers鈥�, as
Drexler calls them, will build everything from computers to cars, molecule by
molecule, from vats of raw materials. In this world, nanobots self-replicate and
self-repair, and since they work in parallel, are speedy and incredibly cheap.
Someday, says Drexler, people will grow everything from plastics to rocket
engines.
Advertisement
But it鈥檚 hard to imagine how to create the first nanomachine, without a
nanomachine to build it. Even though we already have many nifty
nano-components鈥攕ilicon wheels and carbon nanotubes, for
instance鈥攖here is still no reliable way to shuttle them about, position
them exactly where you want and assemble them. That need is not lost on the
proponents of miniaturisation. 鈥淚n the field of nanotechnology, we are getting
good at making objects with neat properties,鈥� says Richard Smalley, a physicist
at Rice University. 鈥淣ow we need to pay attention to how to put them where we
want them.鈥�
This is where Vogel鈥檚 trains chug into the picture. Although she doesn鈥檛 yet
envisage them building computer chips or single-handedly hauling us into
Drexler鈥檚 world, her nanotrains could be the first devices to fill
nanotechnology鈥檚 locomotion gap. 鈥淲hat you need for all these science-fiction
devices is transport,鈥� she says. 鈥淭his may be the first step.鈥�
Of course, nature has always known about the need for transport, and that鈥檚
what drew Vogel to cow brains. Every animal and plant cell contains a
transportation network which carries raw materials, finished products and waste
to their destinations. These highways are long protein rods called microtubules.
Tiny molecular motors race along these microtubules, hauling chemicals from one
end to the other. Nerve fibres, the longest cell extensions in the body, are
stuffed with these highways.
Vogel鈥檚 colleague Jonathon Howard in the biochemistry department at the
University of Washington studies the molecular workings of a motor called
kinesin that is found in the cells of cow brains. Kinesin is a long thin protein
with two stubby 鈥渓egs鈥� at one end and two fat 鈥渉eads鈥� at the other that 鈥渨alks鈥�
along microtubules, one step at a time, carrying its chemical load in a membrane
sack called a vesicle (see Diagram).
One step at a time
Each stride is a puny eight nanometres long. While one leg is bound to the
microtubule, the other swings forward and lands a short way in front. Howard has
found that the power for that swing is provided by the energy in a molecule of
ATP. Both legs of kinesin then stay bound to the microtubule until another ATP
comes along to drive the next step. The result: the kinesin strides along at
about one micrometre per second.
Howard has also shown that each kinesin can exert up to 6 piconewtons of
force as it walks. On the human scale, that鈥檚 measly鈥攔oughly the force
generated by the photons of a laser pointer hitting a projection screen. But on
the nanoscale, it鈥檚 tremendous鈥攅nough to bend a stiff microtubule in half,
for instance. 鈥淭he inside of a cell is a jungle,鈥� Howard says, 鈥渁nd you need a
strong motor to move things through it.
This performance makes kinesin an ideal motor to drive Vogel鈥檚 nanotrains.
For years, people who study cellular motors have made microtubules skitter
around willy-nilly on microscope slides coated with kinesin. Stick the protein鈥檚
heads to the glass, they discovered, and the two stubby legs swing free. They
grab microtubules from solution and start walking them about at random, passing
them from molecule to molecule.
But random motion is of little use, so Vogel has devised a way to create
microscopic tracks to guide her tiny trains. She rubs the surface of a glass
slide with a block of Teflon, a polymer used in nonstick coatings. As the Teflon
scrapes across the slide, rod-shaped polymer molecules shear off onto the glass
where they streak out in fat bundles like fistfuls of spaghetti piled alongside
one another. The result is a pattern of parallel ridges about 25 nanometres
high, separated by about the same distance. Finally, she spreads kinesin
molecules across the slide and drops her trains on top.
Among these diminutive rolling hills, the microtubules no longer move at
random. Instead, snagged by the kinesin, they race off parallel to the grooves.
Vogel has attached a fluorescent dye to her microtubules, so all the action is
visible. 鈥淵ou can really watch them move around the slide with a microscope,鈥�
she says.
The scene through the microscope resembles a bird鈥檚 eye view of a busy
railway goods yard. The glowing microtubule trains move together in parallel or
glide past one another in opposite directions. Once in a while, a microtubule
appears to switch tracks by veering off to the left or right, and then travels
parallel to the other microtubules once more.
To control her trains, Vogel must learn how to constrain them so they always
follow a particular track. This could be tricky. For one thing, when kinesin is
spread on the surface, it can stick in the grooves or onto the Teflon hills on
either side, causing random track switching. Worse, each kinesin is about 75
nanometres long鈥攖hree times longer than the depth of the grooves鈥攕o
the trains may be riding well above the grooves on kinesin posts, like a
miniature monorail. At this height, the trains can easily be pulled out of
alignment by kinesin molecules stuck to the hills on either side, and jump
tracks.
Luckily, Vogel and student John Dennis have found a way to stop her trains
from changing tracks. They altered the concentration of kinesin on the surface
and watched the trains through a microscope. With too little kinesin, they found
the trains have trouble locking on to more than one motor, so they just whirl
around like a propeller on a single kinesin. With too much kinesin, some of the
trains travel diagonally across the tracks, almost as if there were no grooves
at all. As Vogel and Dennis will report in the December issue of
Nanotechnology, only at an intermediate concentration do the trains
actually follow the tracks. They speculate that kinesin molecules prefer to
settle in the grooves. Only when all the grooves are filled do they start
sticking to the ridges. Now that Vogel has learnt how to keep her trains on
track, the next phase can begin: building more complex layouts with curves and
junctions.
Technologies already exist for making single grooves on a nanoscale. Beams of
electrons or X-rays, for instance, can etch ultra-fine patterns in silicon, but
these techniques produce grooves around 100 nanometres wide鈥攆ar too large
to restrict a nanotrain to a single track.
Nanoscale pen
But chemist Chad Mirkin and his colleagues at Northwestern University in
Evanston, Illinois, have a new technique that could solve this problem. They
have learnt how to sketch out nanoscale ridges with a very fine 鈥減en鈥� adapted
from an atomic force microscope (AFM). The nib of this pen is only a few dozen
nanometres across and is held just above the writing surface without touching
it.
When used for microscopy, the AFM tip senses changes in atomic makeup as it
glides above the specimen. But scanning must be carried out in a vacuum, or
moisture in the air condenses at the tip and forms a water bridge between the
tip and the surface. Mirkin took advantage of that bridge for writing. Since
surface tension constrains the water to an area not much bigger than the tip, if
the tip has first been dipped into ink containing small organic molecules, the
water channels it into a fine line like the nib of a fountain pen.
In January, Mirkin reported in Science (vol 283, p 661) that
鈥渄ip-pen鈥� nanolithography could write lines as thin as 30 nanometres across. Now
he says the resolution is down to 10 nanometres, and he expects it to get even
finer. That means the technique could easily create grooves the right width for
microtubule trains by writing two parallel lines just nanometres apart. 鈥淒ip-pen
would be perfect for that,鈥� he says.
Complete layouts, with kinesin-energised tracks laid in loops, figures of
eight or any other pattern, may not be far off. And Vogel has already begun to
imagine how she will equip her trains to do useful work in their molecular
domain.
To assemble anything from an ocean-going liner to a molecular motor, each
component must be brought to a precise location exactly when it is needed. Vogel
and Howard imagine this is how their trains will work鈥攍oading components
at a stockpile, unloading them at the assembly site and then heading back for
more. Nanotrains could drag a molecule past a succession of catalysts and
reagents, for example, and deliver it to wherever the product, whether it鈥檚 a
molecular bearing or a tiny cog, is being built. Other trains could bring other
parts, some of which would have been designed to self-assemble once they get
close enough.
Among the cargo could be nanotubes鈥攔olls of graphite that resemble a
tubular buckyball. Computer chips made of carbon nanotubes would be a thousand
times smaller than current circuits, but the technology to build them doesn鈥檛
yet exist. One solution, says Marvin Cohen, a physicist at the University of
California at Berkeley who is working on ways to build nanocomputers, is to test
random assemblies of nanotubes until you find one that works as a computer
(鈥淲arts and all鈥�, New 杏吧原创, 7 November 1998, p 52).
But to create such a machine by design requires controlled assembly. Vogel鈥檚 molecular railway
fits the bill, he says. 鈥淚f you could control it, it would allow you to grow
肠辞尘辫耻迟别谤蝉.鈥�
But Vogel鈥檚 next step is not nearly so elaborate. First she must prove that
her trains can carry anything at all, let alone nanotubes. So the first cargo
the trains haul will be simple to see: tiny fluorescent beads. To attach them,
Vogel plans to make microtubule trains covered with molecules of biotin鈥攁
vitamin found in egg yolk. Beads coated with the protein avidin would stick
irreversibly to the microtubules, since avidin and biotin bind tightly together.
In a year or so, Vogel hopes to watch these fluorescent beads zip around on her
trains.
By loading different colour beads on trains travelling in opposing
directions, Vogel could also see how close cargoes can be brought together and
whether they can make contact without derailing the train鈥攊mportant
factors when it comes to real assembly. Howard even envisages the energy of
these collisions being used to power synthetic chemistry.
Since the trains have only just left the drawing board, Vogel says it鈥檚 hard
to know what their first practical application might be. It will probably
involve some kind of assembly line. To synthesise a polymer, for example, trains
might move through a series of liquid-filled chambers. In the first, a solution
of monomer A awaits. A single monomer binds to the train, which chugs past a
catalyst that activates the monomer, preparing it for chemical reaction. Next,
trains coming the other way add other monomers onto monomer A and let go, so now
the train is carrying a polymer chain. At the end of the line, the train
deposits its cargo onto a growing sheet of plastic and returns to the first
chamber to take on a new load.
Flash of light
For an assembly line to work, nanotrains need a few more components and
controls. For instance, biotin and avidin aren鈥檛 suitable for cargo that needs
to be unloaded, because once together they are very hard to separate. Instead,
molecular loads will hang on by special bonds that break when illuminated with a
burst of light. And if you want to assemble asymmetrical molecules in a specific
orientation, you could attach the molecules to trains using different types of
linkage, one at each end of the microtubule. These would hold the molecules in
predetermined orientations until assembly begins.
For now, there is no way to make the trains reverse. So getting them back to
the start of the assembly line to reload will probably involve a simple looped
track.
But their speed should be easy to control. Although the top speed will remain
around one micrometre per second, the trains slow down if less ATP is around.
Howard says they could even be brought to a screeching halt with ATP
鈥渕ops鈥濃�� efficient enzymes that consume ATP. Provided there is a steady
supply of ATP, the mops and the motors keep working and the trains keep running.
But when the supply dries up, the mops soak up the excess and everything stops.
A flash of ultraviolet light could refresh the ATP supply by releasing 鈥渃aged鈥�
ATP, a commercially available version of the molecule hooked up to a chemical
blocker by a UV-sensitive bond.
If the nano world is ever to see legions of nanobots transform a soup of
chemicals into a spaceship, for instance, something like a nanotrain will be
needed. For Vogel, it鈥檚 some of the steps in between that are hardest to
imagine. 鈥淚t鈥檚 important to have long-range goals,鈥� says Vogel. 鈥淭he ideas are
fascinating, but what we will really do with this in 20 years is anyone鈥檚
驳耻别蝉蝉.鈥�