EXPECTATIONS are high among the pioneers of the gene revolution. "More will happen in biology in the next 10 years than in the past 50," says Craig Venter, the entrepreneur whose company shared the glory for sequencing the human genome at a White House ceremony on 26 June.

His main rival is just as enthusiastic. "We should be able to uncover the major hereditary contributions to common illnesses like diabetes and mental illness, probably in the next three to five years," says Francis Collins, head of the publicly funded Human Genome Project, which celebrated the completion of its "working draft" at the same White House ceremony. Bill Clinton crowned the event, describing the human genome as "the most wondrous map ever produced by mankind".

Many drugs are ineffective for large numbers of people

Why all the excitement? Because genome research looks set to transform medicine beyond recognition. Modern medicine has hardly advanced since the butchery of Victorian times, according to Peter Goodfellow, a leading geneticist at SmithKline Beecham's pharmaceuticals research centre at Harlow in Essex. "Surgery is still very primitive," he says, and drugs are seriously flawed, too. "There aren't enough of them, the drugs we have don't work that well and there's a significant proportion of patients who don't benefit," he told a recent conference on health care and genetics at the Royal Society in London. In support, he displayed a table showing how few patients respond to today's medicines (see Diagram).

Master plan

What about the medicine of the future? Can the successful sequencing of the human genome usher in a golden age of medicine? What's the pay-off for patients, now that we have the genome laid out before us? And when can we expect revolutionary new treatments that cure the hitherto incurable, without spilling a drop of blood?

What we're talking about is the genetic master plan for assembling and maintaining a human being. All our cells, except red blood cells, carry copies of the master plan. Written with an alphabet of just four letters--the nucleotide bases adenine (A), guanine (G), cytosine (C) and thymine (T)--a single set of the code is some 3.1 billion letters long. There are 24 different chapters, or chromosomes. We get duplicates of 22 of them--one copy from each parent--plus two sex chromosomes. Girls inherit an X chromosome from each parent, whereas boys receive an X from their mother and a Y from their father.

The instructions for assembling human tissue and the cellular machinery that keeps it in good order appear in "sentences" called genes. Faulty genes can cause disease. The haemophilia which cursed Europe's royal families, for example, was caused by a defective gene which messes up the manufacture of a key blood-clotting protein. By decoding the genome scientists expect to discover exactly how genetic misprints cause disease. It should then be possible to produce a drug or therapy to compensate, just as insulin is given to diabetics and clotting factor to haemophiliacs.

Already, researchers working on the Human Genome Project have discovered scores of genes linked with disease (see New Scientist, 20 May, p 14). But the tools are now becoming available to accelerate the haul (see "Enigma variations", p 34).

Finding the gene is just the first step. "You can have all the genes in the world, but what you need is to find the link between them and disease," says Allen Roses, director of genetics at drugs giant Glaxo Wellcome. The roots of diseases like haemophilia, which result from a single faulty gene, are relatively easy to spot by studying family medical histories. But finding the origins of more common killers--heart disease, say, or cancer--is more difficult because in most cases no single gene is to blame. Instead, people inherit labyrinthine patterns of genes which somehow interact to raise the risk of disease. To complicate things further, the risk may only materialise through subtle interplay with the environment, such as breathing polluted air, drinking too much booze or smoking cigarettes.

Despite all these confounding factors, the gene sleuths are pressing ahead, teasing out links between genes and disease by studying real human subjects. One of the best-known efforts is taking place in Iceland, pioneered by the company deCODE Genetics. The company claims that genes are easier to spot in the Icelandic population of around 270,000 because medical records and family histories have been meticulously recorded. They also claim that the population is relatively "pure" because it sprang from a small number of original settlers, although some Icelandic researchers dispute this. But deCODE claims to have made breakthroughs already, announcing last month for example that it has discovered a gene linked with schizophrenia.

Another way to figure out what role defective genes play in diseases is to use what's called a "SNP map" (see "Close cousins"). SNPs, or single nucleotide polymorphisms, are the variants in genes that make each of us a distinct individual. Although 99.9 per cent of our genes are identical to anyone else's, the 3 million SNPs that occur about once every thousand bases account for the remaining 0.1 per cent that makes us unique.

Locating where the most common SNPs lie on each chromosome provides gene sleuths with an invaluable set of signposts for disease-related genes. To this end, a group of 10 pharmaceuticals companies, five academic institutes and the Wellcome Trust medical charity set up the SNP Consortium in April 1999 to compile an SNP map, to be made freely available to all. In September, the consortium published its first map, showing the positions of some 800,000 SNPs. But as with the genome itself, the public SNP effort faces keen competition from Venter. Earlier the same month, Venter's company--Celera Genomics of Rockville, Maryland--announced that it has compiled an SNP map of 2.8 million bases. The Celera map is available only to paying subscribers.

As well as signposting disease genes, SNPs could also provide patients with a more immediate benefit--distinguishing those who can benefit from a particular drug from those who can't. For example, some people might have a variant of a liver enzyme which destroys the drug before it can do any good. If GPs could screen your blood for the crucial SNPs, they could select drugs that are guaranteed to work for you and avoid giving you ones that don't. Such tests could save lives, says Goodfellow. "Adverse drug reactions in 1994 were the fourth highest cause of death in the US." Companies such as Affymetrix of Santa Clara, California, aim to producing "gene chips" capable of reading thousands of SNPs at once from a single patient. They hope these will make SNP testing affordable.

Meanwhile, Venter believes the most direct way to uncover a gene's function is to see what its counterpart does in other creatures. "Comparative genomics will be the single most important tool for analysing genes," he says.

Among the creatures that have already been sequenced or are now being sequenced are fruit flies, nematodes, yeasts, pufferfish, zebrafish, frogs and a host of disease-causing microorganisms. By disabling individual genes through mutation and seeing what happens to the creature, geneticists can work out the functions of the animal genes, and then check whether there are defects in people's genes that tally with those in the sick animal.

More ambitious are the bids to sequence the genome of the mouse, which closely matches the human genome. Speaking in Britain in July, Venter boasted to a conference on genetics in Birmingham that Celera will have sequenced the mouse by December (New Scientist, 22 July, p 14). This has triggered another public-private race. A consortium that includes the Wellcome Trust and the US National Institutes of Health responded last month with a £39 million drive to achieve the same goal by next February. Venter promptly announced that 95 per cent is already done.

Meanwhile, the new flavour of the month is proteins--each one the product of a different gene recipe. Genes may carry the master plan, but it's the proteins that make up the stuff of life--some 200 types of living tissue in humans as well as hundreds of thousands of enzymes that keep our tissue ticking over.

Grand venture

Never one to rest on his laurels, Venter has launched an ambitious project to automate the exploration of the human "proteome", the protein equivalent of the genome. "At best, our gene code will tell us the increased risk of disease," he told delegates in Birmingham when announcing his grand venture. "It's only by understanding protein function that we can truly understand and predict medical outcomes."

Proteomics pioneers hope to spot disease-causing proteins through a type of identity parade which compares the protein profiles of healthy and diseased tissue (see "High in protein", p 38). Venter announced in Birmingham that he is linking up with 44,000 cancer surgeons in the US to chart protein profiles. Similar public initiatives have been set up, such as the Cancer Genome Project at the Sanger Centre near Cambridge. But protein profilers have their work cut out, because some genes create their proteins in several sections, which can then be spliced together in any order, so a single gene can generate many different proteins. Proteins can also have their functions altered or be switched on and off simply by adorning them with sugars or phosphate groups.

Illustration: Nick Lowndes

But combined study of genes and proteins will take researchers closer to the ultimate goal of the genome project--understanding the molecular causes of disease and devising drugs which intervene at just the right point. A particular faulty protein might, for example, be the crucial switch that sets off a catastrophic chain reaction leading to cancer. Once researchers understand the protein "wiring diagrams" in cells, they can spot the faulty wiring and short-circuiting which causes disease. Make a drug that hits the protein "fallout" from the chain reaction and you're only treating the symptoms. Hit the "master protein" or the gene that makes it and you might be able to stop the cancer in its tracks, or prevent it in the first place (see "Know your enemy", p 46).

Even after researchers have worked out the biology and where to intervene to prevent or cure disease, they still need a drug to do the business. Fortunately, companies have worked out ingenious methods for making and screening millions of potential drugs simultaneously, selecting which ones bind to the critical protein to block it, activate it or do whatever it takes to cure the disease. For instance, NeoGenesis, a company in Cambridge, Massachusetts, has devised a system for testing novel chemicals on key disease proteins at a rate of 300,000 a day.

Then when you've found a chemical that does the trick, you have to make sure it isn't too toxic to give to patients. Lab tests on colonies of living cells give a preliminary answer. But prospective drugs are then tested in animals, before they are finally tested on people. There are plenty of drugs that look good in early trials but fail abysmally at the final fence--when tested against a dummy or existing drugs.

It takes at least a decade for new drugs to clear these hurdles. "The completion of the full sequence and the SNP map are important milestones, but only signify the end of the beginning," says Arthur Holden, chairman of the SNP Consortium. The first pay-off--the potential for personalising treatments with existing drugs--is already in sight. New wonder drugs to treat currently incurable diseases will have to wait longer.

Still, the genome, the SNP map and the proteome will provide a mighty boost in the fight against disease. For the first time, all the culprits causing human disease--whether genes or proteins--will have been rounded up in one place.Expect progress to accelerate. At the end of his talk in Birmingham, Venter pointed out that he spent 10 years cloning and working out the DNA sequence of the human receptor for adrenalin. Now, using the equipment at Celera, the same task would take 15 seconds. Similarly, Roses says that linking a gene to part of the genome would have taken five years in 1995 and 6 months in 1999. By 2001, with the SNP map and enough data from patients, it could take as little as a week.

"My dream is the eradication of common diseases," says Holden. Society may have to be patient, and wrestle with some weighty moral and ethical questions to reach these goals, but most people would sooner face that than the surgical butchers of Victorian days.

Other Genome features available Beyond the Genome: Now that the genome revolution has begun, New Scientist takes a look at how it will change our lives Close cousins: It's vanishingly small. A mere 0.1 per cent is the difference between your genome and mine. But those tiny fragments could help us get to grips with killer diseases, says Kathryn Brown