/ 14 May 1999

Flesh and chips

Cyborgs have feelings too. When California neurosurgeon Gary Heit activated one of the implants he had just put into a human brain, the man burst into tears. Heit’s patient at Stanford university neurosurgery clinic was crying because, for the first time in 15 years, he was able to button up his own shirt.

A debilitating tremor, similar to that caused by Parkinson’s disease, meant this 79-year-old man had been unable to feed or clothe himself. Silicon technology and biological material – hard wiring and soft flesh – can now be combined, but so far, the results have remained unmistakably human.

Actor Christopher Reeve, paralysed from the neck down in a riding accident in 1995, could be enabled to walk again with technology already available, Heit says. “If somebody gave me sufficient funds to pull a research group together I could get Christopher Reeve to walk in two years. I could probably make him stumble right now.”

Heit says it is simply lack of money that holds the field back from giving greater benefit to those debilitated by lack of brain function. A brain implant for an injured human being can cost more than $20 000, which he believes is a trivial sum compared with the way it revolutionises patients’ lives.

Cases like Reeve’s would require an array of implanted high-speed microprocessors to analyse brain activity and translate those impulses into electrical commands. Those commands could be passed to a set of electrodes attached to the nerves that would stimulate muscles to perform a walking action.

A few kilometres from the Stanford campus, a biotech spin-off company is about to begin commercialising another development in the silicon-biology interface, invented by researchers at the University of Bangor.

Aura Diagnostics of Mountain View, California will develop a laboratory-on-a- chip capable of isolating cancer cells in tiny blood samples. The process, which uses electrical signals to manipulate biological molecules, should enable early detection of cancers. It could also provide simple, quick and accurate tests for meningitis, and a means for regenerating damaged tissues and organs.

It is one of many similar innovations now moving from research prototype into commercialisation: the Institute for Cancer Research unveiled its own “cancer chip” last month.

The Bangor researchers were unable to find British backing. As was always the case with the computer chip industry, commercial exploitation of the silicon-biotechnology interface is still largely confined to American soil. Dawn may reach the United States a few hours later, but the future always seems to arrive there first.

Seeing the lucrative potential of this particular future, US firms that made fortunes through information technology are now looking to apply their expertise to biology.

Motorola has just begun work on a project with the US Department of Energy’s Argonne national laboratory and Packard Instruments, a Connecticut-based pharmaceutical instrument company. The aim is a mass production technique for “biochips” – thumb-size wafers that carry out information processing on the chemicals of life.

Unravelling the human genetic code will provide ways of diagnosing and treating genetic diseases and cancers, as well as giving us an instruction manual for the assembly of the human body. But this is a complex business, and painfully slow when done by standard laboratory means.

By combining biochips with robots and computers the Argonne researchers claim they will be able to find one genetic variation among three billion DNA data bits in a matter of minutes. Conventional methods take days.

Like the microchip, the biochip will be mass-produced by photolithography techniques, a big operational change for conventional biotech companies.

`From our point of view, these biochips are a huge dimension shift,” says Rolan Papen, business development manager for Packard Instruments. “You have to start thinking of clean rooms and miniaturised components. The companies that have a lot of experience with this are people like Motorola.”

Making a biochip involves laying down a forest of DNA material onto a glass substrate. There can be as many as 400 000 DNA strands on a chip, and each one is designed to trap – and thus detect – a certain type of DNA sequence.

The four chemical blocks that make up DNA – generally referred to as the bases A, T, C and G – only bind together in certain ways, as with Ts and Cs with Gs.

The DNA strands attached to the biochips contain carefully designed sequences of these bases attached to the glass substrate. The DNA to be tested is broken up into fragments which are tagged with a fluorescent chemical and then washed over the chip. The way these fragments bond with the DNA on the biochip, shown up as the chip is scanned with a laser pulse, allows researchers to infer their genetic makeup.

The ultra-fast processing power of the DNA chips is set to change the field of medical diagnostics, researchers believe. Tiny blood samples will be processed within minutes to test for the presence of mutated genes that signpost dangers of cancer, multiple sclerosis or Alzheimer’s. Bacteria and viruses such as HIV will be instantly identified, and drug-resistant strains will also be spotted.

“The chip technology is going to bring diagnosis from the laboratories to the physician’s office,” says Tuan Vo-Dinh, group leader of the life sciences division of the US Department of Energy’s Oak Ridge national laboratory.

Vo-Dinh’s research group has just licensed the commercialisation of its own silicon biochip, an integrated HIV, tuberculosis and cancer detector. The matchbox-size combined chip is made from CMOS silicon and also contains logic circuits for data processing.

Getting the prototype to market will definitely involve the co-operation of companies in the microprocessor industry, Vo- Dinh says. But, he says, it needn’t take long. “I can see them reaching commercialisation in the next one to three years.”

Use of fused silicon and biology once seemed like science fiction, but it is likely to become commonplace within just a few years.

Packard’s Rolan Papen says. “The day when a customer will walk into a general store, buy a biochip, perform a test and take some medication as a result is a little bit away.”