The enemies within: The origins of the Cancer Genome Project 1/3/02. By Giles Newton The Cancer Genome Project is systematically searching for genes involved in cancer.

"About one in three people will get cancer over their lifetime, and about one in five will die of it," says Professor Michael Stratton, who leads the project. "Critically, this huge burden on human health results from abnormalities in human DNA. With the human genome sequencing efforts coming to fruition, we can use this vast resource to find the genes involved in cancer and the changes to the normal sequence that are causing such a huge burden of disease."

To find the cancer genes, the Trust awarded £10 million to Professor Stratton and Richard Wooster of the Institute of Cancer Research, to set up a Cancer Genome Project at the Sanger Institute and to 'test drive' this approach to the discovery of disease genes.

Genes and cancer

During everyday life, human cells acquire changes to their DNA as a result of exposure to the environment - chemicals, radiation or viruses - or as a result of errors in the enzymatic machinery that copies DNA during cell division. Most of these changes are put right by the cell's repair mechanisms, but if mistakes occur in a subset of genes known as oncogenes or cancer genes, the cell in which those changes occur can start to behave abnormally. The cell may continue dividing when it should be stopping and instead of occupying its normal position it may infiltrate the tissues around it and ultimately spread to other parts of the body.

Like all cancers, breast cancer is caused by mutations in human DNA.

"The recognition that cancer is a disease of DNA has been with us for 40 years, and cancer researchers have made the finding of cancer genes one of their main priorities," says Professor Stratton. "The main strategy used has been to localise those genes to a small area through genetic methods - for example by looking at families with many cases of cancer and finding the areas of the genome that confer susceptibility - or by looking down a microscope for break-points in the chromosomes of cancer cells. These approaches have been very successful, albeit time consuming, and about 100 cancer genes have been found."

But many cancer genes are thought to be undiscovered. Most of the 100 known genes have been found in the relatively rare leukaemias and lymphomas, which account for less than 10 per cent of all human cancer; for the common adult epithelial cancers of the breast, colon, prostate, lung and ovary, which account for 80 per cent of the cancer burden, only about 30 genes are known.

"It has also become clear that our current strategies for finding cancer genes are running out of steam. It is becoming more difficult to map cancer genes to small, well-defined areas of the genome," says Professor Stratton. "And some cancer genes cannot be mapped at all in that sort of way. For example, several genes are activated in cancers due to a single amino acid substitution in the protein; such genes do not leave an imprint on the genome that can be seen by studying families or by looking down the microscope."

Enter the human genome sequence

A new approach was needed, and the human genome sequence offered an exciting alternative. "Richard Wooster and I had been working on new strategies to find cancer genes," says Professor Stratton. "It was obvious that the advent of the human genome sequence was going to revolutionise this area of research and so we wanted to find ways to use the whole genome sequence and systematically to go through all 70 000-100 000 genes asking the question 'is this gene mutated in cancers?'."

"Five different types of genetic abnormality can occur in cancer cells," says Professor Stratton. "We're starting off by searching for two types, but all five come under our remit and must ultimately be addressed." The first task will be a search for homozygous deletions - where both copies of a gene are lost - in approximately 1000 cancer cell lines. This project will systematically test for the presence of every gene in each cell line. The second search will look for small abnormalities in the genomes of cancer cells: base-pair substitutions, or small insertions or deletions. Ultimately it will screen every gene in the genome, in a 'search set' of 48 cancers mainly from common adult epithelial cancers.

"Once we have found the cancer-causing genes, we will need to clarify which genes are abnormal in which cancers. We also need to investigate which cancers have many abnormalities and which have few abnormalities," says Professor Stratton. The team therefore plans to go on to screen 2000 cancers of all types - childhood, adult, epithelial and soft tissue - for abnormalities in these genes.

"It's very difficult to say at the outset how many genes are involved or how many we will find. We have speculated that there may be about 400 cancer genes, but I am happy to be proved wrong."

Links

The Cancer Genome Project at the Wellcome Trust Sanger Institute