|Unstable Genomes and Cancer|
|Eleven thousand mutations|
|By Bijal P. Trivedi
May 5, 2000
Early stage cancer cells can be riddled with as many as 11,000 mutations sprinkled all over their genome. Sometimes the mutation rate is even higher, and in late stage tumors the number of mutations can be as high as 100,000. This explosion of mutations is called genomic instability because the damage affects the cell's entire complement of DNA, or genome.
For decades, scientists have been studying what genetic abnormalities must occur for a cell to become cancerous. They have found mutations where genes have been duplicated, relocated, inactivated and deleted. They have also identified a set of about 30 genes that are mutated in a broad range of cancers. But until the late 1990's many scientists believed that mutations in only a few of these genes are required to trigger cancer. Now new evidence suggests that genomic havoc precedes a cancerous state.
Lawrence Loeb proposed in 1974 that cancer is the result of a heavily mutated genome, but he lacked the molecular tools to prove the point. One reason the field didn't move forward is that, until recently, there was no good animal model to test the idea, says Loeb, of the University of Washington in Seattle. Another hindrance was the absence of an assay to find the random mutations that were occurring in a genome with three and a half billion components.
Now the field is beginning to move. Garth Anderson and colleagues at the Roswell Park Cancer Institute in Buffalo, NY, have devised a technique to find certain types of random mutations, while other researchers have engineered experimental models that exhibit instability.
A small number of genes, such as APC, Ras, and p53, which are commonly mutated in many cancers, were each hailed at one time or another to be the key to cancer, and later were collectively thought to be the backbone of cancer progression. But it appears these may be a few mutations among thousands, and more key players remain to be found.
"11,000 mutations are pretty striking," says Joe Gray of the University of California, San Francisco. But which, and how many, of these mutations are significant remains unknown, he adds. "It's only within the past 5 years that people are beginning to appreciate the magnitude of instability."
When Anderson's team looked at late-stage cancer cells from 58 patients with non-hereditary cases of colon cancer, they found about 11,000 mutations per tumor cell. The researchers then turned to pre-cancerous growths to determine whether these mutations caused cancer or were the result.
Cells from colon polypstissue growths that are prone to become tumorshad many of the same 11,000 mutations as the late-stage tumors. This suggests that wide-scale DNA damage begins early and contributes to tumor development, making it likely that the large number of mutations is the cause rather than the result of cancer.
Sentinels of the Genome and Mutator Mice
When genes involved in replicating or repairing DNA are mutated, a cell with an unstable genome is often the result. A couple of recent reports demonstrate the point.
André Nussenzweig and colleagues at the National Institutes of Health focused on the Ku80 gene that is used throughout the body to repair breaks in DNA. Breaking and rejoining occurs naturally in many parts of the genome and is especially important in immune cells where regions of the DNA are constantly being broken and reshuffled to generate novel antibodies.
In mice lacking the Ku80 gene the DNA was cut in many places but could not reconnect to the correct piece of DNA; the results were cells with unstable genomes. Some chromosomes lost large chunks of DNA, which other chromosomes gained. Many cells carried small bits of DNA and the wrong number of chromosomes, some of which were two or three different chromosomes fused together.
On their own, these unstable cells rarely cause cancer because the p53 gene triggers death in cells when DNA damage is beyond repair. When Nussenzweig created mice without Ku80 and p53, the mice got B-cell lymphomas before three months of age. When these sentinel genes are removed, instability can lead to cancer, says Nussenzweig. The work is presented in a recent issue of Nature.
Brad Preston and Robert Goldsby, both of the University of Utah, decided to trigger instability via a different route. Instead of mutating a DNA repair gene they attacked genes that copy and replicate DNA before the cell divides. It makes intuitive sense that if DNA polymerase, which replicates DNA, doesn't "proofread" its work, then there are going to be lots of errors, or mutations, says Preston.
Preston and Goldsby created mice in which the DNA polymerase was unable to "proofread" the DNA it copied. Preston refers to the rodents as "mutator mice" because their DNA has a devastatingly high mutation rate. By ten months of age 40-50 percent of the mice had cancer.
"[Our work shows that] changing the mutation rate can cause cancer, but whether it is absolutely necessary remains to be seen," says Preston. Preston presented these results at the annual Environmental Mutagen Society Meeting, in New Orleans, Louisiana, in early April.
Implications for Cancer Therapies
Slowing or preventing genomic instability could provide a new approach to treating cancer according to Anderson and Loeb.
Anderson, along with many others, believes that measuring the extent of genomic instability may also be a useful tool in cancer diagnosis. If genomic instability occurs early in tumor progression, it may be an early warning signal that the genome is mutating at an accelerated pace.
Currently, Anderson's lab is focusing on five locations around the genome that are mutated early in tumor progression, quite probably because of instability. One possibility is that enzymes that repair mutations in normal cells have been lost or may no longer function leading to an unstable genome. A breakdown in DNA repair would cause damaged DNA to accumulate, producing the large number of mutations the research team found. Anderson is examining the mutation sites for recurring patterns that could shed light on the cause of damage.
. . .