Tumor survey unearths wealth of mutants
Cataloguing cancer genes may pay dividends.
A hunt through thousands of human genes has turned up nearly 200 that are altered in breast and colon cancer. These genes might be useful for diagnosing cancer or as new targets for drugs.
The findings suggest that a US$1.5-billion initiative funded by the US government to decode the 'cancer genome' could throw up useful leads and goes some way towards appeasing the project's critics.
Cancer arises when cells rack up mutations in a number of their genes, and begin to divide uncontrollably. Researchers have already identified some of the genes involved in this process by cherry-picking promising candidates but there are thought to be many, many more.
In the latest study, researchers searched for culprits by determining the genetic sequence of some 13,000 genes found in 11 breast tumours and 11 colorectal tumours that had been preserved for study. They then looked for differences in the genes between cancerous and normal tissues, and cross-checked the result with an additional bank of tumours from 24 breast or colorectal cancers.
The trawl unearthed a total of 189 genes that were mutated in the tumours and are suspected to be a cause of cancer; the majority of these had not been implicated in cancer before.
The study confirms that cancer is a fiendishly fickle enemy. The team found that breast and colon tumours harbour almost completely different mutations in fact, only two mutated genes were shared between them. Cancers in other tissues might also be driven by a different spectrum of mutations.
In addition, the team found that no two tumours are exactly alike. All in all, the researchers estimate that a typical breast tumour carries mutations in more than 100 genes. Some 20 of these might be involved in causing the cancer, they say. Less than half of these are likely to be found in another breast tumour. The study is published in Science1.
The results add weight to the idea that battling cancer is going to be a long and difficult task, as each person might need a different, tailored combination of drugs to combat the wayward cancer genes that are fuelling their disease.
But there is some hope. One of the lead authors on the study, Victor Velculescu of the Sidney Kimmel Comprehensive Cancer Center in Baltimore, Maryland, points out that many of the mutated genes fall into groups with a common function, such as sending signals within a cell. So it is possible that drugs that interfere with these processes might work on many different cancers. "Simpler themes emerge within the complexity," he says.
The study serves as a forerunner of the Cancer Genome Atlas, a vast project that aims to find all the genetic changes that occur in cancers. A $100-million pilot project, funded by the US National Human Genome Research Institute and the National Cancer Institute, is just starting up; but some scientists have been critical of the whole approach, saying that it might not unearth mutations that occur in a significant number of cancers, and so might be of little use.
The new study, done independently of the atlas, allays some of those concerns because it shows that at least some such mutations do occur in a significant number of cancers, even if they are not found in all cancers. "It's a good initial study. I'm glad that [the Atlas team] will find something interesting because I'd hate all that money to go to waste," says cancer researcher Stephen Elledge of Harvard Medical School in Cambridge, Massachusetts, who has been critical of the Cancer Genome Atlas in the past.
Elledge says that he still has qualms over the amount of money to be channelled into the project. He argues that some of it might be better spent on studies aimed more directly at finding new drugs - such as blocking genes one by one in order to find those that can stop cancer cells dividing. "I still think it is better to have a more balanced approach that includes functional studies," he says.
Visit our newsblog to read and post comments about this story.
- Sjoblom T., et al. Science, 10.1126/science.1133427 (2006).