Spinning a Good Tale
by Alexander Gelfand
The Economist
Dec. 4, 2008
BIOTECHNOLOGISTS have long dreamed of creating a "lab on a chip" that would pack the power of a full-scale analytical laboratory into an object as small and as easy to use as the hand-held scanners familiar to fans of science fiction. Such a device might detect biological weapons, run genetic tests or sniff out contaminants. Staff at clinics could use it to screen people for infectious diseases. Police could perform on-the-spot drug tests; paramedics, roadside diagnoses.
Many designs have been proposed for such a device, but none has really taken off. The latest, though, sounds promising. It uses a quantum-mechanical effect called giant magnetoresistance (GMR), which is also the basis of a computer's hard drive. And prototypes made in laboratories in Europe and America have indeed been able to detect everything from deadly toxins to illegal drugs and markers of disease.
Giant magnetoresistance relies on devices called spin valves. These are made by interleaving thin sheets of magnetic and non-magnetic metals to form a sandwich composed of layers mere nanometres (billionths of a metre) thick. If a nanoscale sandwich is exposed to a magnetic field, the quantum spin of its electrons, and thus its electrical resistance, will change in a way that is easily detectable—which is why they are used in the heads of hard-disk readers.
In 1998, however, it occurred to David Baselt of the United States' Naval Research Laboratory in Washington, DC, that spin valves might also make excellent biosensors. Biological materials themselves are not usually magnetic, but they are often chemically specific. To Dr Baselt this suggested that tiny magnetic particles might be attached to molecules using either antibodies (which will bond to proteins, sugars and so on) or single-stranded DNA (which will bond to a complementary DNA strand to form the famous double helix). To search for a target molecule, then, all that would be needed would be to sprinkle a sample thought to contain it with magnetic nanoparticles coated with the appropriate antibody or DNA and then run it over a spin valve.
A simple idea. It has, though, taken Dr Baselt and his colleagues, along with researchers in several other organisations, ten years to turn his insight into a practical technology. Now that they have, things are about to go mainstream.
Though the details of the prototypes differ, the principles are similar. Both the spin valves and the paraphernalia of channels needed to feed a sample to them are built onto silicon wafers using the same techniques employed to make microprocessors. The wafer is then cut up into chips.
If the antibody method is being used, the tester deposits a droplet of sample fluid that has been seeded with magnetic particles coated with the appropriate antibodies onto the chip's surface. The target molecules will thus stick to the particles. The sample and the particles then drift along the channels towards spin valves that have been prepared with different antibodies to the same target molecules. Any particle with a target molecule on its surface will stick to the spin valve. There, its magnetic field is detected and the signal sent to a display. The DNA method works slightly differently, in that the sample and the particles are treated with chemicals that cause DNA to stick promiscuously to the particles. In this case it is the spin valves that are covered with complementary strands to the target, so that when the particles are washed over them, those that have target strands on their surfaces will stick. In either case, each spin valve reacts to only one sort of molecule, so the chips can search simultaneously for as many types of molecules as they have spin valves.
Such chips are extraordinarily sensitive. Prototypes built by researchers at the Naval Research Laboratory, Stanford University and Philips, a Dutch electrical-goods firm, can detect a substance that is present at a concentration of less than one part in a million billion. Liesbet Lagae, who works at IMEC/MCP, a microelectronics laboratory in Belgium, and Paulo Freitas, of INESC, a similar organisation in Portugal, claim even more striking results. Their experiments can sometimes detect individual molecules. In particular, this sensitivity makes the new chips better at searching for strands of DNA. At the moment, such searches require the number of DNA molecules in a sample to be multiplied using a trick called the polymerase chain reaction before there is a decent chance of detecting the target.
The GMR-based method also offers advantages other than sensitivity. In fluorescent assays, for example, the target molecules are tagged with glowing labels that can be seen through a microscope. But blood contains many chemicals that emit light naturally, resulting in a lot of background noise. In addition, searching for more than one kind of biomolecule using standard methods often requires large samples and a series of assays. A GMR chip can do the job in a single pass with a sample as small as a single microlitre.
Nor are these advantages merely theoretical. Shan Wang of Stanford has used a GMR chip smaller than a postage stamp to detect carcinoma-embryonic antigen (a sign of colon and rectal cancer) and human papilloma virus, the main cause of cervical cancer. And, at the Naval Research Laboratory, Cy Tamanaha has had similar success detecting ricin, staphylococcal enterotoxin B and other potential bioweapons. According to Dr Tamanaha, the laboratory has developed a portable, battery-powered unit the size of a shoebox, and has licensed its technology to Seahawk Biosystems, a firm based in Austin, Texas, for food-testing and environmental applications.
Others researchers use magnetism not only as a way of detecting the target, but also as a way of carrying it to the spin valve. In a paper in the Journal of Immunological Methods, Wendy Dittmer of Philips claims to have achieved a 100,000-fold increase in detection speed by hurrying magnetically tagged molecules of parathyroid hormone towards the spin valve with an electromagnet instead of allowing them to diffuse there. Dr Lagae and Dr Freitas have used similar techniques to find marker molecules for breast cancer and genes linked to cystic fibrosis. Philips plans to introduce a hand-held drug-testing device in 2009, in partnership with Cozart, a British medical-diagnostics firm. This magnetically enhanced breathalyser, designed for roadside use by police, will scan suspects' saliva for drugs such as cannabis, cocaine and heroin. According to Menno Prins, who led the effort to develop Philips's magnetic biosensors, the technology can suss out morphine "in less than one minute".
There may be an even faster method, however—and one that avoids fiddling with electromagnets. Marc Porter, at the University of Utah, has brought nanomagnets and GMR sensors together with nothing but a simple credit-card-style swipe. In a pair of papers to be published soon in Analytical Chemistry, Dr Porter describes how he and his graduate students ran a slender glass stick larded with magnetic nanoparticles through a GMR reader. Dr Porter believes that an automated reader could scan a sample stick in a second, and he envisages a day when pharmacies will keep such readers on premises, with customers bringing in personal "wellness cards" preloaded with bodily fluids and obtain their diagnoses without having to sit forever in a doctor's waiting room.
Copyright ©2008 Alexander Gelfand