SURELY it is every bench scientist’s dream. A computer-controlled laboratory
that can perform a quarter of a million experiments in parallel, at the touch of
a button. All you have to do is sit back and let the results roll in.
Far-fetched? Not if you ask Gianni Medoro, chief scientist at a small company
called Silicon Biosystems in Bologna, Italy. He and his colleagues are on their
way to building just such a lab, but you won’t find it full of white-coated
scientists or robots. This research facility is inside a wafer-thin chamber atop
a silicon chip. Here, the team can hold and manipulate thousands of separate
living cells suspended in an electric field, shifting them around as easily as
if they were counters on a draughtboard.
Not much use on its own, you might say. But equip this tiny chamber with
sophisticated biosensors and you can perform tests on individual cells. It could
revolutionise medical research—and might one day even save your life.
Advertisement
For years, biotechnologists have been working to create chips packed with
sensors that can diagnose disease in minutes from a single drop of
blood—the idea behind so-called lab-on-a-chip research. The devices
developed so far work fine on controlled samples, but they stumble when they are
faced with real ones. Blood, for example, contains dozens of kinds of cell, and
millions of different molecules. This complex soup confuses the sensors and
generates unreliable results—a big problem when misdiagnosis could
kill.
Silicon Biosystems’ cell manipulator could be the lifeline lab-on-a-chip
technology has been looking for. Build an array of biochemical sensors into the
device and it could pick out the cells to be examined and subject them to a
battery of tests, while ignoring all the irrelevant junk around them. It could
also offer reliable results almost instantly—perfect for making a speedy
diagnosis in hospital or at the doctor’s surgery. It could even offer an
automated “while-u-wait” service at your local pharmacy.
Silicon Biosystems’ chip is based on a phenomenon called dielectrophoresis.
If you place a living cell in a non-uniform, alternating electric field, charges
are induced on the cell’s membrane. The field then exerts a force on these
charges, pushing the cell around. By controlling the frequency at which the
field alternates, it is possible to alter how the cell is charged, and the
direction in which it moves. Pick the right field and frequency, and you can
shunt the cell about whichever way you choose.
This technique is already the basis of simple biochips that can separate one
kind of cell from another
(91av, 1 March 1997, p 22). But in
1999, Günter Fuhr and his colleagues at Humboldt University in Berlin
showed that as well as just sifting cells, dielectrophoresis could be used to
trap individual cells and move them around.
They positioned tiny electrodes around a small chamber and filled it with a
suspension of living cells. Then they switched on a high-frequency electric
field. At one particular frequency, the fields from the electrodes interfered to
create regions of weak electric field in the chamber. The cells quickly moved
towards these field “minima” and became trapped. The minima were acting like
tiny electric cages. The researchers then altered the electric fields to move
the cages smoothly about the chamber, taking their trapped cells with them.
Although Fuhr has proved the technique works, his system has some problems,
according to Silicon Biosystems chief executive Nicolo Manaresi. The
micrometre-sized electrodes have to be precision-machined and carefully aligned
in three dimensions. This makes the system extremely fiddly to build.
The answer, Manaresi and his colleagues decided, was to design a silicon chip
to do the job. This allows them to make use of the powerful manufacturing
technologies developed by the chip industry. For example, the lithographic
techniques devised to etch electronic circuits would make it possible to put up
to a million electrodes on a single chip.
Silicon Biosystems’ prototype contains a hollow chamber that holds just a few
cubic millimetres of solution. Across the floor of the chamber are long, thin
electrodes 50 micrometres wide. The top of the chamber is sealed with a glass
lid on which a transparent layer of conducting indium tin oxide has been
deposited to act as the top electrode. The chamber is mounted on a circuit board
and the electrode voltages are controlled by computer
(see Diagram).
In tests, the researchers found they could trap and move around tiny
polystyrene beads just micrometres across. “We were also able to create
dielectrophoretic cages which can trap living [yeast] cells,” says Manaresi.
Most impressive of all, they found that they could separate a mixture of
yeast cells and polystyrene beads, and control each group individually. To start
with, a mixture of the particles sat on the bottom of the chamber. When the team
applied the first signal, the beads and cells rose together and were trapped in
a row of cages in the middle of the chamber. By switching the frequency of the
signal, they could release the cells so that they drifted to the floor, while
keeping hold of the beads. Then they shunted all the beads into one cage. When
the team switched frequency again, the cells jumped up and were trapped.
The device gave the researchers complete control. Had this been a real
biochemical experiment, with beads coated in drug molecules, the cells and
molecules could have been brought together, made to interact, and then separated
and collected for analysis.
However, the prototype’s large cages make it difficult to trap and move
single cells. To overcome this, their next device will be a chip with about a
million electrodes arranged on the bottom in a two-dimensional array. Each
electrode will be a few square micrometres in area, and the chamber will be just
80 micrometres deep—about the thickness of a human hair. With at least
four electrodes per cage, this array will be able to produce up to a quarter of
a million independently controllable cages.
Will this device really work as planned? “Yes, I’m quite sure about that,”
says Masao Washizu of Kyoto University in Japan. Washizu, an expert in
dielectrophoresis, says the Silicon Biosystems design is better at trapping and
moving cells than anything else he’s seen.
Manaresi’s “grab and drag” technique is all very well, but it’s of little use
if you can’t tell what you’re handling. That’s where Marco Tartagni comes in.
While at the University of Bologna, he patented the world’s first
microelectronic fingerprint sensor that works by monitoring the changes in
capacitance between pairs of electrodes when a finger is pressed against them.
With Tartagni’s help, the team has designed a way of sensing when a cell is
trapped. And they hope his system will eventually be able to tell what kind of
cell has been caught.
The idea is simple. When a cell is trapped in a cage, it slightly alters the
capacitance between the top and bottom electrodes. The change depends on the
type of cell, so if you could measure it, you could learn to tell a skin cell
from a muscle cell, for example.
In practice, however, sensing such tiny changes in capacitance requires some
fancy electronic footwork. While the signals that create the cages use
alternating voltages, a steady voltage is needed to measure capacitance. To see
if they’ve trapped a cell, the researchers have to switch off the alternating
field and apply a steady field for just long enough to take a
measurement—but not so long that the trapped cells can escape.
The researchers are confident they can do this. Once they have made the
device, they can begin to characterise different types of cell—including
blood cells, liver cells and skin cells—according to their capacitance
measurements. Then they can begin to combine their chip with arrays of
biochemical sensors—technology that already exists, thanks to
lab-on-a-chip research.
Integrating cell manipulation and biochemical sensing in silicon is a very
powerful concept, says Jon Cooper, a professor of bioelectronics and
bioengineering at the University of Glasgow. The chip offers a way to analyse
thousands of cells in parallel. Better still, it should be a simple matter to
program the chip to move its contents about in whatever way you want.
Milan Mrksich, a biochemist at the University of Chicago, Illinois, is also
enthusiastic. Imagine you’re searching for a drug candidate that kills cancer
cells without damaging normal tissue, he says. If you have cages holding cancer
cells in one spot, liver cells in another and skin cells somewhere else, you
could treat the array with various compounds to see which ones affect only the
cancer cells. “I think this could provide an important technology for drug
discovery,” Mrksich says.
Dying or alive
Carol Dahl of the National Cancer Institute in Bethesda, Maryland, is more
interested in the prospects it offers for basic research. Currently, it’s
difficult for biochemists to study individual tumour cells, and to differentiate
between cells that are dying and those that are not. “If this technology could
be used to allow more comprehensive molecular analysis of individual cells, it
could be very useful in the discovery of the molecular basis of cancer,” she
says.
But Manaresi admits that there are still challenges to overcome. For
instance, not every cell type will cooperate: many cells have to be attached to
a protein scaffold in order to survive and function normally. So we simply don’t
know whether it’s possible to investigate the properties of some tissue cells
when they are floating free in solution, says Mrksich. And body fluids such as
blood contain salts that make the fluid conduct electricity, potentially
interfering with the connections.
Nonetheless, manipulating cells with dielectrophoresis certainly works.
Researchers elsewhere have already shown that the herpes viruses can be trapped
and moved around in dielectrophoretic cages. Others have separated cancerous
cells from healthy cells in mouse blood and have even grabbed proteins and latex
beads that are just nanometres in diameter—thousands of times smaller than
the cells that Manaresi uses.
So could Manaresi’s mighty machine end up juggling individual viruses or even
molecules? “I don’t see why not,” he says. “Just turn up the field and off you
.”