Old laboratory mice develop substantially fewer and less-aggressive lung tumors than younger animals in a new study led by Stanford University researchers. The discovery flies in the face of established dogma that holds that cancer risk increases with age, but it dovetails with what's seen in very elderly people, in whom cancer risk appears to either level off or even decline with age.
"It's a striking finding," associate professor of genetics and of pathology Monte Winslow , PhD, said. "We would expect that older animals would get more and worse cancers, but that's not at all what the study found. So, what is it about the molecular changes associated with aging that suppress cancer?"
Despite the intriguing hint that aging may have a silver lining, it's been difficult to study the phenomenon thoroughly in the laboratory. In the absence of much data, it's not been possible to know how a cancer patient's age might affect their responsiveness to various treatment options.
"Aging is a systemic change to your body, but most cancer studies in mice are conducted on younger animals," said former graduate student Emily Shuldiner, PhD. "When we introduced the same lung cancer-causing mutations in young and old mice, the young mice develop more, faster-growing tumors."
Shuldiner is the lead author of the study , which was published Nov. 4 in Nature Aging. The senior authors of the paper are Winslow and professor of biology Dmitri Petrov , PhD.
Factoring in the age factor
It makes intuitive sense that cancer risk increases with age. Each time our cells divide, there's a chance of fumbling the genetic ball and introducing mutations into our DNA that can hamstring a cell's ability to respond appropriately to signals that tell them when to divide and when to chill.
For most of our lives, that model seems to hold true: When plotted against age, cancer incidence in the general population rises sharply beginning around 50 years and peaks around 70 to 80 years. But after about 85, the curve plateaus and even drops. It's not been clear whether that shift is due to a decrease in screening, and thus in diagnosis, or perhaps a kind of natural selection — maybe people who are predisposed to longevity also have immune systems better equipped to eliminate developing cancers?
But the results in the mice suggest a deeper biological cause.
"The standard model of cancer is, with age you accumulate bad things in the form of mutations," said Petrov, the Michelle and Kevin Douglas Professor in the School of Humanities and Sciences. "And when you collect enough bad things, cancer happens. After a certain age it should be almost inevitable, right? But that's not what we see; after a certain point aging appears instead to be a generic form of cancer suppression."
Other changes also accumulate with age: For example, the patterns of chemical tags on our DNA called methyl groups that help regulate which genes are expressed and when they are expressed are altered. And the genome generally becomes more structurally unstable and vulnerable to breakage. Segments of DNA in our cells' energy factories can duplicate and reintegrate into the genome, creating stretches of repeating sequences like identical boxcars on a train track.
Among all that genetic mayhem, however, some of these changes hamstring cancer development in a way that could be exploited for new therapies. But identifying this in animals is extremely time consuming.
Shuldiner tackled the problem head on. The mice in her study were genetically engineered to develop fluorescently tagged lung cancers when the animals were treated with an inhaled gene delivery system. But to compare tumor formation between young (four to six months) and old (20 to 21 months) mice, she first had to wait nearly two years for the animals to age. (The average lifespan of a laboratory mouse is about two years.)
Once the animals were sufficiently long in the tooth, Shuldiner induced lung cancer formation. Fifteen weeks later, the amount of cancer in the lungs of the young mice — measured by lung weight and fluorescent imaging — was about three times that of the older mice. The young mice also had about three times as many tumors, and these tumors were significantly larger than those found in old animals.
"In every way we could measure, the younger animals had worse cancers," Shuldiner said.
She then investigated the effect of inactivating 25 tumor-suppressor genes in the animals before triggering cancer development. These genes make proteins that block the development of cancers, and many are involved in processes associated with normal aging.
When it's better to be old
The impact of inactivating any of several tumor suppressor genes were lessened in old mice. That is, although tumor incidence increased in the animals with inactivated tumor suppressors regardless of age, the effect was greater in younger mice. But inactivating one tumor suppressor gene in particular, PTEN, had a much greater impact than the others.
"PTEN inactivation stood out as having a much stronger effect in young mice," Shuldiner said. "It suggests that the effect of any given mutation, or the efficacy of cancer therapies targeted at specific mutations, might be different in young versus old people."
Shuldiner also investigated the gene expression patterns in cancer cells from old animals with either active or inactive PTEN.
"We found that patterns known to be associated with aging were still present in the cancer cells from old mice," Shuldiner said. "This was not an obvious finding because cancer cells are rapidly dividing. It was interesting to see those signatures of aging remained. However, in old animals in which PTEN was inactivated, these aging signatures in the cancer cells were much less pronounced. They looked as young as the PTEN-deficient cancer cells from young animals, which was very surprising."
The study is the first to conclusively show that aging represses tumor initiation and growth and that it changes the impact of inactivation of tumor suppressor genes like PTEN. It also describes how signatures of aging remain in old cancer cells, even though they are dividing rapidly. The findings illustrate the importance of developing new cellular and animal models of cancer that incorporate the effects of aging to develop new therapies.
"We develop these animal models of cancer with an eye to developing new treatments for patients," Winslow said. "But for this to work, the models have to be correct. And this study suggests that models using young animals may not be accurately reflecting important aging-related changes."
"The implications of this story could be huge," Petrov said. "Maybe aging has a beneficial side to it that we could harness for better therapies."
The study was funded by the National Institutes of Health (grants R01 CA234349, R01 CA230025, U01 AG077922, P30 CA124435 and F99/K00 CA2344962), the Canadian Institute of Health Research, the Stanford Cancer Institute, the American Cancer Society and the University of California's Tobacco-Related Disease Research Program.
Winslow is a member of the Stanford Cancer Institute , the Wu Tsai Neurosciences Institute and Bio-X . Petrov is a member of the Stanford Cancer Institute, the Maternal and Child Health Research Institute , and Bio-X. He is also a Chan Zuckerberg Biohub investigator.
