The results, reported in the July 27 on-line version of Nature Genetics, support the emerging viewpoint that common variant forms of some genes can contribute to cancer susceptibility, particularly when a number of these variant genes are co-inherited and therefore can act in concert.
Traditionally, advances in understanding cancer predisposition have involved discoveries of rare, inherited, full-fledged genetic mutations that dramatically increase the risk of cancer, such as the brca1 and brca2 genes in breast cancer. But scientists have come to recognize that such "high penetrance" genes account for only 5-10 percent of cancers.
Evidence from mouse models and human epidemiologic studies suggests that susceptibility to most cancers results from numerous low penetrance polymorphisms acting together to disrupt a cell's regulatory controls. In themselves, these subtle variations are not harmful. A mouse – or human – that inherits a mildly more active form of a certain regulatory gene such as myc or ras won't necessarily have a problem. But the evidence suggests that if either inherits a number of overly active regulatory genes that act synergistically, cancer is more likely to develop.
"It's like the components of a lock – it's the combination of these polymorphisms that makes a substantial difference in susceptibility," says senior author Allan Balmain, PhD, UCSF professor of biochemistry and Barbara Bass Bakar Endowed Chair in Cancer Genetics at the UCSF Comprehensive Cancer Center. "We're trying to identify the components of the lock. The Aurora2 story is the first component we've identified."
While much evidence has suggested that cancer susceptibility genes exist in humans, identifying them has proven daunting. In the current study, the researchers used a novel approach involving both mouse models of cancer susceptibility and resistance, and human tumors to systematically move in on the Aurora2 gene, and to identify the culprit variation of the gene, known as Ile31.
Scientists already knew that an abnormal number of copies of the Aurora2 gene, known as a kinase, were found in more than half of colon tumors, indicating the gene's possible role in destabilizing the cell. But they have not known whether it actually contributes to the development of cancer, or is a by-product of cancer.
The new finding, says Balmain, should heighten scientific interest in exploring whether an inhibitor of the gene, already being developed by pharmaceutical companies, helps to prevent the recurrence of colon tumors and to treat existing cancers.
The discovery of the Ile31 variation's apparent impact on cancer susceptibility should also broaden interest in the gene as a potential therapeutic target for many other cancers, including those of the skin, breast and prostate, where the researchers suspect varying degrees of activity.
Moreover, he says, the investigational approach used by the researchers, which allowed them to probe the genome at an usual levels of speed and detail, could be used to identify other human susceptibility genes.
The Aurora2 gene codes for a protein that plays a key role in ensuring that the correct number of chromosomes -- two for each of the 23 pairs -- is passed to each daughter cell during cell division. But the Ile31 variation makes the gene slightly more active than normal, causing some daughter cells to inherit extra copies of chromosomes. The abnormal and haphazard increase or decrease in the number of chromosomes, a form of aneuploidy, disrupts the growth control of the cell.
While the impact of genomic instability on cancer development remains hotly debated, Balmain and many others believe that it incites a spiraling chain reaction of unregulated behavior that can lead to the accumulation of mutations in key regulatory genes. When enough mutations accumulate, the cell moves into replication overdrive, one of the hallmarks of cancer.
The Ile31 genetic variation is likely to be a particularly important factor in cancers in which an environmental stimulus – almost always a contributing factor in cancer -- has prompted the cell into a state of rapid cell proliferation, says Balmain, as cell division will be occurring at an increased rate. Given this, he says, colon and skin cancers are particularly likely targets, as a diet high in saturated fat can prompt colon cells to proliferate, and ultraviolet radiation from the sun makes skin cells proliferate.
Notably, almost a quarter of people inherit the Ile31 variant of the Aurora2 gene, but most do not develop cancer. Given this, says lead author Amanda Eward-Toland, PhD, a postdoctoral fellow in Balmain's lab, "the variant is a classic low penetrance susceptibility gene. This is just the beginning of this field, where we are trying to establish what the networks are that make us susceptible."
The evidence suggests that multiple low penetrance genes can confer a dramatic increase in susceptibility to cancer. In women without a family history of breast cancer, says Balmain, there can be a 40 to 50 fold increase in susceptibility in a woman because of the particular combination of low penetrance genes she inherited from her parents.
Ultimately, he says, "If scientists could identify the network through which the Aurora2 variant works, clinicians potentially could treat patients with drugs that target several of the genes involved. In this way, we could dismantle the network that controls the progressive growth of the cancer cell. That's the critical point we're aiming for in this field. It's network analysis."
The current study is almost as notable for the strategy used to identify the Aurora2 gene as for the discovery itself. The researchers examined mice that had been bred to inherit susceptibility genes that caused the growth of tumors and, in an approach known as haplotyping, moved in on a relatively small region of DNA that they identified as likely to contain susceptibility genes. (While scientist would normally have to examine a region of 10 to 20 million bases pairs of genes, they were able to refine their search to 1 to 2 million base pairs, a ten-fold reduction in the size of the piece of DNA.)
After identifying 10 possible susceptibility genes in the mice, the researchers turned to human tumors to see if they could identify similar genes there. They identified three. Aurora2 looked the most promising. They then sequenced DNA samples from about twenty different human individuals, identifying the subtle variation now known as Ile31. The researchers examined this variant in tumor samples and cells in culture and determined it was more active.
Co-authors of the study were Paraskevi Briassouli, of the Breakthrough Breast Cancer Research Centre, Institute of Cancer Research, London; John P. de Koning, of the University Medical Center, Utrecht, The Netherlands and of the UCSF Comprehensive Cancer Center; Jian-Hua Mao and Jinwei Yuan, of the UCSF Comprehensive Cancer Center; Florence Chan, of the Cancer Research UK Centre for Cancer Therapeutics, Sutton, UK; Lucy MacCarthy-Morrogh, of the Breakthrough Breast Cancer Research Centre; Bruce A.J. Ponder, of the Cancer Research Campaign Human Cancer Genetics Group, Strangeways Research Laboratories, UK; Hiroki Nagase, of Roswell Park Cancer Center, Buffalo, New York; John Burns, of Newcastle University, Newcastle, England; Sarah West, of Newcastle University, Newcastle, England; and Spiros Linardopoulos, of the Breakthrough Breast Cancer Research Centre and of the Cancer Research UK Centre for Cancer Therapeutics, The Institute of Cancer Research Haddow's Laboratories, UK.
The study was funded mainly by a National Cancer Institute Mouse Models of Human Cancer Consortium Grant, the UCSF Prostate SPORE Program, Cancer Research UK and the Breakthrough Breast Cancer Fund.
Journal
Nature Genetics