快猫短视频

FOR dozens of families in Lanarkshire last Christmas will be remembered not as a happy time, but with deep sadness. They were at the centre of Britain鈥檚 worst outbreak of Escherichia coli food poisoning, which struck down more than 300 people and tragically killed 18 of them. The first symptoms were spotted in mid-November, but by the time the offending organism was identified as the deadly strain known as 0157:H7, more than 50 people were in hospital and one man had died. The delay in identification was nobody鈥檚 fault-the technology simply could not deliver any sooner. Whether anybody died as a direct result of the delay will never be known, but there is no doubt that faster diagnosis could have speeded up treatment, minimised suffering and allowed more spot checks on food to identify the source of contamination.

Pinning down the strain of pathogens such as E. coli is desperately time-consuming. The bacteria are first grown on agar to produce enough for testing and then infected with viruses that attack specific strains. If one virus kills the bacteria, then microbiologists have a good idea of what they鈥檙e up against. The identity is confirmed using enzymes to cut the bacterium鈥檚 DNA and copy the fragments, which are then dragged by an electric field through a gel. This process, called electrophoresis, leaves a characteristic pattern of fragments-a 鈥渇ingerprint鈥� that is compared with the fingerprints of known strains.

鈥淭his is the state of the art at the moment,鈥� says Hugh Pennington, professor of medical microbiology at the University of Aberdeen, who analysed the samples from the Lanarkshire outbreak. 鈥淲ith people breathing down our necks we can get results in about four to five days.鈥� He adds ruefully: 鈥淎nything that can speed up this process would be welcome.鈥�

His prayers are close to being answered. Research groups worldwide are shrinking labs onto tiny squares of silicon or glass to speed up the detection and identification of bacteria. These 鈥渓abs on chips鈥� don鈥檛 just shrink conventional analytical methods, they are a proving ground for a whole raft of novel techniques for manipulating and identifying microorganisms and controlling chemical reactions. And it鈥檚 not just microbiologists who stand to benefit. Chemists are getting to grips with pumping substances through tubes only micrometres wide, mixing them in vats smaller than a fingernail and heating them with needle-like elements. Miniaturisation promises them unprecedented control over their reactions. Some chemists claim that in the long term these tiny labs could revolutionise their science in much the same way that microprocessors transformed electronics 25 years ago, and all our lives since.

One of the research teams closest to putting a whole lab on a chip is led by Ron Pethig of the Institute of Molecular and Biomolecular Electronics at the University of Wales, Bangor. The group鈥檚 cell-handling chip depends critically on the ability to move cells around by turning them into tiny dipoles with the aid of an external electric field. The principle is analagous to rubbing a pen and attracting a small piece of paper to it. The rubbing leaves the pen with a positive charge which distorts the distribution of charges in the piece of paper, turning it into a dipole. The pen鈥檚 field repels the paper鈥檚 positive charge and attracts the negative. The net effect is to pull the negative end of the paper closer to the pen.

Pethig does not use a constant electric field, however, but one that oscillates at between 100 hertz and 10 megahertz. This range allows him to tune the frequency so that he can move different types of cell. A 100 kilohertz field, for example, will move E. coli one way and Micrococcus lysodeikticus the other. This happens because several structures associated with cells respond to electric fields: liquid molecules immediately around cells adopt a characteristic pattern of charges, as does their outer surfaces, membranes and internal structures. And since different cell types vary in one or all of these aspects, they respond differently to different field intensities and frequencies.

Pethig uses different layouts of electrodes to move cells around a chip. An electric field travelling along silicon electrodes laid like railway sleepers, for example, will push bacteria along like a wave carries surfers (see Diagram). He can also separate different types of microorganisms at forks in these electrode tracks-it鈥檚 all a matter of applying different fields to the two prongs of the fork. This technique works well for separating, say, different species of bacteria, but it cannot always separate one strain of a species from another. In this case, something needs to be added that will modify the electrical behaviour of one strain says Julian Burt, a member of the Bangor team. The researchers achieve this by adding polystyrene beads 6 micrometres across that are coated with antibodies specific to one of the strains.

Spin your microbe

If any members of the target strain are present, they are eventually swept into a round chamber enclosed by electrodes that generate a rotating electric field. It鈥檚 here that for the past three years, Pethig and his colleagues have been spinning cells and cataloguing how they behave. The researchers have spun everything from baker鈥檚 yeast to a host of bacteria, and from their motion can tell not only which microorganisms they have collected but also whether they are alive or dead.鈥滻f you have more than a certain number of live ones you could be in trouble,鈥� says Burt.

If small numbers of the target cells are present, it鈥檚 possible to count them under a microscope, or sometimes with the naked eye. Bacteria tagged with polystyrene beads spin very differently from other bacteria and can easily be identified, says Pethig. A camera and some simple image processing software should automate the counting process, but for higher numbers a better method may be to shine light through the chamber and see how much is absorbed. This should give a measure of the concentration of bacteria in the chamber, says Burt.

By putting all the elements together, Pethig and his colleagues hope to create a desktop tester with the area of a Pentium chip and a few millimetres high. They hope the device will be so sensitive and specific that bacteria will not always have to be grown on agar gel before testing. The device should also catch particles as small as viruses.

Such a system would be attractive not only to public health officials but also to commercial companies. One of the big players in the trend to miniaturisation is the water industry, and much of Pethig鈥檚 work has been done in collaboration with Severn Trent Water. Parasites such as giardia and cryptosporidium are all too common in drinking water. Consuming just 30 cryptosporidia is enough to trigger a long-lasting bout of diarrhoea. The parasites are resistant to standard disinfectants and at just 5 micrometres across can slip through filters used to trap other particles.

Quicker test results would give water companies the opportunity to take more samples, giving them a better chance to catch contaminants before someone drinks them. As Issy Cafoor, head of research at Yorkshire Water, says: 鈥淨uite often we get results well after the water has been past the point we sampled.鈥� A water supplier is as likely to find out about contamination from angry or ill customers as from the company labs.

Pethig鈥檚 chip should identify cryptosporidium and giardia in less time and in smaller numbers than is possible using conventional means. Aspects of the chip are being tested by Genera Technology, a subsidiary of Severn Trent Water now, but a full lab on a chip is still a couple of years away at least.

Policing water

Not surprisingly, the ambitions of the Bangor group are shared by others. In Britain, for example, Yorkshire Water is working with Siemens and Southampton and Luton universities to create ceramic chips for detection work. They want to build autonomous, disposable sensors to police the water supply, radioing messages to headquarters as soon as they find something suspicious.

But labs on chips are not confined to juggling microorganisms. Another group eager to speed up testing are microbiologists who analyse DNA. In recent years, the polymerase chain reaction, which allows researchers to make millions of copies of a single strand of DNA, has revolutionised microbiology. And while much of the PCR process has been automated, it is still a time-consuming business.

PCR works by using one strand of DNA as a template for rebuilding its complementary strand. First, the DNA under investigation is heated to 100 掳C so that the double helix splits apart. Next, it is cooled to 55 掳C and short DNA stretches, or primers, are attached to the two ends of the target sequence. The mixture is then heated to 72 掳C along with the four chemical bases that make up DNA and the enzyme DNA polymerase. This encourages the bases to bind to the primers and to one another to regenerate the complementary strand. The mixture is then heated to 100 掳C and the whole process repeated again and again.

Typically, researchers and technicians carry out this elegant technique in a thermal block cycler, which is designed to heat and cool up to 50 samples at once. But there is a problem with this approach. Because of its large thermal mass, few samples stay at the correct temperatures for the correct lengths of time. 鈥淕enerally speaking, PCR still needs relatively large amounts of DNA because current thermal cycling techniques are inefficient,鈥� says Andrew de Mello, who is shortly to become the Zeneca lecturer in analytical science at Imperial College, London.

Before he returned to Britain last year, de Mello was a member of a team at the University of California, Berkeley that designed a chip to carry out both PCR and electrophoresis-the first of its kind in the world. The PCR takes place in a 2 centimetre-high silicon reaction vessel that sits on top of a flat electrophoresis wafer.

Small is beautiful

Although relatively crude, the chip slashed a single PCR cycle from between 2 and 6 minutes with a thermal block to less than 30 seconds. Enzymes are added to cut up the DNA and, when the PCR is complete, electric fields draw the fragments into gel-filled capillaries beneath the reaction vessel. Normally, electrophoresis takes hours to separate the fragments. But on the chip, it鈥檚 all over in less than two minutes.

鈥淲e use very high electric fields to get rapid separation,鈥� says de Mello. Normally, the heat generated by a high voltage would destroy the gel. 鈥淚n a narrow capillary you can dissipate the heat very efficiently, so you can use high electric fields for electrophoresis,鈥� he says. Using the Berkeley chip, the team identified a human blood protein from scratch in 20 minutes and Salmonella DNA in less than 45 minutes.

The notion of being able to do PCR on a chip is now attracting big bucks. The US government is pumping millions into developing the concept, chiefly as a means to speed up and automate sequencing of the human genome. Also, the US Army is starting to use a PCR chip in a battlefield kit for analysing the DNA of soldiers killed in action.

Perhaps the biggest impact of miniaturisation could be in analytical chemistry. Making chemicals react is not always as straightforward as school textbooks would have you believe. Mix chemicals for too short a time and a reaction can be left incomplete, too long and the reaction may go too far, creating unwanted products. Even the wrong temperature can produce unwelcome results. 鈥淎nalytical chemistry before now was all about buying an instrument and doing quality control,鈥� says Andreas Manz, head of the analytical chemistry department at Imperial College, London.

Complete control

For the past couple of years, Manz has been the driving force behind a 鈥渕icro total analysis system鈥� (Micro TAS). Such a device would have capillary tubes and reaction chambers, sensors for measuring properties such as temperature and pH, and control and data-collection electronics all integrated on a silicon chip. Being able to govern precisely how tiny amounts of substances move on the chip should give chemists unprecedented control over their reactions. Manz claims that if such devices are realised they will revolutionise analytical chemistry, speed up drug discovery and help to tease out the detail of complex chemical reactions.

At the Central Research Laboratory of Thorn-EMI in Hayes, Middlesex, Ian Robins and his colleagues are in the early stages of designing a Micro-TAS system. They are studying ways to control the movement of chemicals through capillaries just 100 micrometres wide. As with the Berkeley chip, fluids can be moved simply by applying a voltage between the two ends of a channel. The researchers are also investigating the impact of surface tension-the dominant force at this scale-though it is too early to tell if it will be a help or a hindrance.

鈥淲e do not mix the liquids,鈥� says Robins, 鈥淲e get them together in fine channels, bring them together and the reaction completes itself as fast as chemical kinetics allows it.鈥� The resulting compounds should be much purer than those obtained from mixing chemicals in a test tube. Some reactions have two or three stages and one of the advantages of working at the micrometre scale is that it is easy to cool a compound quickly or move it swiftly to another chamber ready for the next stage.

As well as revitalising current research techniques, labs on chips could open up totally new areas. Pethig and his colleagues, for example, have put human cells through their paces on the electrode tracks and noticed that some do not rotate as fast as their neighbours. Cancerous cells, it turns out, spin differently from healthy cells in an electrical field. The difference stems from changed conductivity of the cell membrane, says Burt, but the precise reason is still unclear. The Bangor team hopes that this difference could be useful as an early test for cancer.

Stiff cells

In a similar vein, Mark Tracey at the University of Hertfordshire and Michael Rampling from Imperial College, London have found that squeezing red blood cells through channels etched in silicon could help to detect diseases such as cirrhosis, cancer and diabetes. Here, a small proportion of red blood cells can become stiffer than normal. This also happens before the diseases are detectable by conventional means.

The stumbling block to using this phenomenon for any kind of test has been spotting the stiff cells. Tracey and Rampling hope that their channels-each about 4 micrometres across or roughly half the width of a typical red cell-will solve this problem. With a camera and some image processing software, they hope to automate the process of identifying when cells have trouble fitting through the channels. For the time being, the technique is still being refined in the laboratory, but the researchers hope that it could eventually become a valuable clinical test.

While there is growing excitement about labs on chips, there are also clear limits to their applications. In some cases, the chips handle only picolitres of fluid, so it is unlikely that they will play a big role in industrial chemical plants, for example.

But even before labs on chips can be used, they have to be made. And building the right structures out of silicon is turning out to be a challenge in itself. For many of the tiny labs, capillaries and pits are about 100 micrometres wide, which is large when you consider that electronics manufacturers are starting to make components just 0.35 micrometres across. But for Tracey, this relatively large size creates its own problems. Pits have to be 鈥渄ug鈥� through several silicon films and keeping the edges smooth is a problem that has yet to be overcome. Blood and bacterial cells latch onto surface irregularities and clog up devices. They could also burst or get damaged and so give false readings in tests.

Smooth surfaces are also essential on Pethig鈥檚 chip: to create uniform electric fields his electrodes must have sharp edges. But he has an extra problem. The optimal size for the electrodes is 10 times the size of the cells they are shuffling around. For cryptosporidium, which is about 5 micrometres across, this is not a problem. But to make electrodes capable of moving viruses-which can be just 10 nanometres across-Pethig and his team will need to push the limits of lithography.

Perhaps the biggest hurdle facing all the designers of labs on chips is integrating all the components on one device. Most of the miniature labs built so far require large pieces of equipment for at least part of the process. Pethig鈥檚 team needs a microscope to detect microorganisms on its chip, for example. The US Army鈥檚 battlefield PCR chip is part of a system that fills a suitcase, while the Berkeley PCR chip relies on an argon laser to make strands of DNA fluoresce.

Solutions do exist. Pethig鈥檚 team could measure light passing through its rotational chamber using optical fibres, for example, or build the entire chamber on top of a light-sensitive charge-coupled device. De Mello is now trying to cut out the lasers on PCR chips by replacing them with thin-film light-emitting diodes. Integrating these solutions is the next key challenge, says de Mello.

Imagine

He likens the situation to the microelectronics industry in the late 1950s and early 1960s when discrete components had been made in silicon, but not combined. 鈥淚magine taking a pipette and placing a drop of liquid on the chip and the chip does the rest-DNA amplification, separation and measurement. And the result comes out the other end,鈥� he says. 鈥淭hat鈥檚 what we鈥檙e after.鈥�