New Ideas on the Progression of Cancer

High mutation rates in cancer cells can be caused by non-DNA damaging stresses

Friday, January 19, 2001

print | email |

Duke University Medical Center researchers report for the first time that long-term genetic instability in cancer cells can be induced both by stresses that break DNA and those thought not to damage genetic material.

The stress-induced persistent genetic instability they have described is likely to play an important role in carcinogenesis - in the progression of cancer even if not in the initiation of it, the researchers said. The results also might help explain the large number of mutations seen in cancer cells, they added.

If other stresses, such as chemotherapy drugs or steroids, are eventually shown to cause similar effects, the results could upset the prevailing view of the mechanism behind cancers developing resistance to therapy or becoming more aggressive, the scientists said.

"In general, it was thought that if a cell suffers a mutation and survives, the mutation is fixed within one or two generations," said lead author Chuan-Yuan Li, assistant professor of radiation oncology at Duke Comprehensive Cancer Center. "We're reporting a new phenomenon - that non-DNA-damaging stresses can cause persistent genetic instability. The frequency of mutations seen in this study is orders of magnitude higher than previously reported."

The results are published in the Jan. 15 issue of Cancer Research, the journal of the American Association for Cancer Research. The study was funded by the National Cancer Institute, the Duke Department of Radiation Oncology, the Komen Foundation for Breast Cancer Research, and a Duke Comprehensive Cancer Center fellowship.

Li and his colleagues compared untreated mouse cancer cells to those exposed to one of five severe stresses, which were designed to kill more than 90 percent of the original cells. Each stress caused spontaneous deletion of a marker gene and alteration of naturally occurring repetitive DNA sequences in the genome of surviving cells and their never-exposed progeny. The stresses tested were ionizing radiation, hydrogen peroxide, high temperature, nutrient starvation, and growing the cells in mice, which creates multiple stresses.

What If . . .

There are mechanisms to explain the number of mutations found in only a minority of cancer cells, the researchers said. In cells that have lost the ability to repair their DNA or those with inherited "cancer genes," for example, the mutation rate jumps to perhaps one mutation per 1,000 cell divisions, up from one per million for "normal" cells.

But because most cancer cells have normal DNA repair abilities, and most people with cancer don't have a recognized hereditary predisposition to developing the disease, repair deficiencies and inherited predisposition aren't enough.

"Basically, if developing cancer depended on acquiring the right set of mutations by chance alone, no one would ever get cancer," said co-author Mark Dewhirst, professor of radiation oncology. "Clearly, people get cancer and there's been no convincing way to explain how the right mutations could accumulate with enough frequency to lead to the incidence of cancer that is observed."

The researchers' finding that severe amounts of some non-DNA-damaging stresses can induce long-term genetic mutability might help explain the numerous mutations seen in cancer cells, but it raises the question of whether the newly described process could be involved in how cancer cells develop resistance to chemotherapy drugs or hormonal treatment.

Developing resistance to cancer treatments is fairly common; prostate cancers eventually become insensitive to hormonal therapy, for example. Scientists generally believe that resistance develops because a few cancer cells in the original tumor might be insensitive to treatment. Thus when all the sensitive cells are killed, only the resistant ones remain, according to the conventional wisdom.

However, if common treatments can induce long-term genetic instability similar to that seen in the study, resistance might "evolve" through accumulation of mutations after treatment, rather than from the initial survival of already resistant cells, the researchers said. Much more research is needed to know for sure, they added.

Li's future studies will examine other stresses and other cell types, and will also explore the possibility of an apparent threshold effect, as reduced stress exposure led to lower long-term genetic instability. The researchers will also try to determine how the stresses caused instability.

"We already know a great deal about how radiation affects normal cells and cancerous cells," said Li. "But we know nothing about the mechanism by which the non-DNA-damaging stresses were able to create persistent genetic instability."

The researchers used large amounts of ionizing radiation and hydrogen peroxide, which, like many other stresses known to damage DNA, create highly reactive oxygen atoms, or free radicals, that cause breaks in DNA. The other stresses used weren't known to have any effect on genetic material.

The Study

In order to evaluate the stresses' effects on DNA, the mouse cancer cells were engineered to express a visual marker - green fluorescence protein (GFP) from a species of jellyfish - that makes the cells glow green when exposed to blue light.

"Because each cell has just one copy of the GFP gene inserted into its genome, if that gene is mutated or damaged, that cell becomes dark," explained Li. "It's a very convenient and direct way to examine the genetic stability."

The scientists also looked for general instability in the cells' DNA sequence, in case the inserted GFP gene was hypersensitive. By examining naturally occurring repetitive DNA sequences called "minisatellites" that are scattered throughout the genome, the researchers proved that the stresses affected native portions of the genome as well as the inserted marker gene.

After the cells were exposed to a stress, the scientists selected ones that still fluoresced green, indicating that the GFP gene was still working, and allowed them to grow under standard laboratory conditions. Much to their surprise, roughly 10 percent of the "daughter" cell colonies several generations after exposure contained a mix of light and dark cells, reflecting new loss of GFP in cells never directly exposed to stress, they reported.

The majority of daughter cell colonies from exposed cells were uniformly bright, as were all cell colonies of cells whose "parents" were not exposed to any stress, indicating no loss of GFP. A few daughter cell colonies of exposed cells were uniformly dark, reflecting complete loss of GFP. The mixed, or chimeric, cell colonies showed that new mutations were still occurring up to 23 generations after stress exposure.

Other co-authors are John Little of the Harvard School of Public Health; Kang Hu, Wen Zhang and Li Zhang, all of Duke; and Qian Huang, now at the People's Hospital, Shanghai.

Written by Joanna Downer.