Matt Ridley

themselves.

If the cell cannot be stopped, we call the result cancer.

But usually, it can be stopped. The problem of cancerous mutiny is so old that in all large bodied animals the cells are equipped with an elaborate series of switches designed to induce the cell to commit suicide if it should find itself turning cancerous. The most famous and important of these switches, in fact possibly the most talked about of all human genes since its discovery in 1979, is TP53, which lies on the short arm of chromosome 17. This chapter tells the remarkable story of cancer, through the eyes of a gene whose principal job is to prevent it.

When Richard Nixon declared war on cancer in 1971, scientists did not even know what the enemy was, beyond the obvious fact that it was an excessive growth of tissue. Most cancer was plainly neither infectious nor inherited. The conventional wisdom was that cancer was not a single form of disease at all, but a collection of diverse disorders induced by a multiplicity of causes, most of them external. Chimney sweeps 'caught' scrotal cancer from coal tar; X-ray technicians and Hiroshima survivors contracted leukaemia from radiation; smokers 'caught' lung cancer from cigarette smoke and shipyard workers 'caught' the same affliction from asbestos fibres.

There might be no common thread, but if there was it probably involved a failure of the immune system to suppress tumours. So went the conventional wisdom.

Two rival lines of research were, however, beginning to produce new insights that would lead to a revolution in the understanding of cancer. The first was the discovery in the 1960s by Bruce Ames in California that many chemicals and radiations that caused cancer, such as coal tar and X-rays, had one crucial thing in common: they were very good at damaging D N A . Ames glimpsed the possibility that cancer was a disease of the genes.

2 3 4 G E N O M E

The second breakthrough had begun much earlier. In 1909, Peyton Rous had proved that a chicken with a form of cancer called sarcoma could pass the disease to a healthy chicken. His work was largely ignored, since there seemed so little evidence that cancer was contagious. But in the 1960s, a whole string of animal cancer viruses, or oncoviruses, were discovered, beginning with the Rous sarcoma virus itself. Rous was eventually given the Nobel prize at the age of eighty-six in recognition of his prescience. Human oncoviruses soon followed and it became apparent that whole classes of cancer, such as cervical cancer, were indeed caused partly by viral infection.

Putting the Rous sarcoma virus through the gene-sequencer revealed that it carried a special cancer-causing gene, now known as src. Other such 'oncogenes' soon followed from other oncoviruses.

Like Ames, the virologists were beginning to realise that cancer was a disease of genes. In 1975 the world of cancer research was turned upside down by the discovery that src was not a viral gene at all. It was a gene that we all possessed, chicken, mouse and human, too. The Rous sarcoma virus had stolen its oncogene from one of its hosts.

More conventional scientists were reluctant to accept that cancer was a genetic disease: after all, except in rare cases, cancer was not inherited. What they were forgetting was that genes are not confined to the germline; they also function during an organism's lifetime in every other organ. A genetic disease within an organ of the body, but not in the reproductive cells, could still be a genetic disease. By 1979, D N A taken from three kinds of tumour had been used to induce cancerous growth in mouse cells, thus proving that genes alone could cause cancer.

It was obvious from the start what kinds of genes oncogenes would turn out to be - genes that encourage cells to grow. Our cells possess such genes so that we can grow in the womb and in childhood, and so that we can heal wounds in later life. But it is vital that they are switched off most of the time; if they are jammed on, the result can be disastrous. With 100 trillion body cells, and a fairly rapid turnover, there are plenty of opportunities for oncogenes to be jammed on during a lifetime, even without the encouragement D E A T H 2 3 5

of mutation-causing cigarette smoke or sunlight. Fortunately, however, the body possesses genes whose job is to detect excessive growth and shut it down. These genes, discovered first in the mid-1980s by Henry Harris at Oxford, are known as tumour-suppressor genes. Tumour suppressors are the opposite of oncogenes. Whereas oncogenes cause cancer if they are jammed on, tumour-suppressor genes cause cancer if they are jammed off.

They do their job by various means, the most prominent of which is to arrest a cell at a certain point in its cycle of growth and division, then release it from arrest only if it has all its papers in order, so to speak. To progress beyond this stage, therefore, a tumour must contain a cell that has both a jammed-on oncogene and a jammed-off tumour-suppressor gene. That is unlikely enough, but it is not the end of the matter. To escape and grow uncontrollably, the tumour must now pass by an even more determined checkpoint, manned by a gene that detects abnormal behaviour in a cell and issues an instruction to different genes to dismantle the cell from the inside: to commit suicide. This is TP53.

When TP53 was first discovered, by David Lane in Dundee in 1979, it was thought to be an oncogene, but it was later recognised to be a tumour suppressor. Lane and his colleague Peter Hall were discussing TP53 in a pub one day in 1992 when Hall offered his arm as a guinea pig for testing if TP53 was a tumour suppressor. Getting permission to perform an animal test would take months, but an experiment on a human volunteer could be done right away. Hall repeatedly scarred a small part of his arm with radiation and Lane took biopsies over the succeeding two weeks. They showed a dramatic rise in the level of p53, the protein manufactured from TP53, following the radiation damage, clear evidence that the gene responded to cancer-causing damage. Lane has gone on to develop p53 as a potential cancer cure in clinical trials; the first human volunteers will be taking the drug as this book is being published.

Indeed, so rapidly has cancer research in Dundee grown that p53

is now bidding to be the third most famous product of the small Scottish city on the Tay estuary, after jute and marmalade.

2 3 6 G E N O M E

Mutation in the TP53 gene is almost the defining feature of a lethal cancer; in fifty- five per cent of all human cancers, TP53 is broken. The proportion rises to over ninety per cent among lung cancers. People born with one faulty version of TP53 out of the two they inherit, have a ninety-five per cent chance of getting cancer, and usually at an early age. Take, as an example, colorectal cancer.

This cancer begins with a mutation that breaks a tumour-suppressor gene called APC. If the developing polyp then suffers a second mutation jamming on an oncogene called RAS, it develops into a so-called 'adenoma'. If it then suffers a third mutation breaking another, unidentified tumour-suppressor gene, the adenoma grows into a more serious tumour. And now comes the danger of a fourth mutation, in the TP53 gene, which turns the tumour into a full carcinoma. Similar multi-hit models apply to other kinds of cancer, with TP53 often coming last.