A Closer Look
Tackling the Complexity of Genes and Environments in Cancer
DNA sculpture by Miroslaw Stuzik (Photo by Tomasz Gasoir)
With few exceptions, cancer and other common diseases are thought to be a result of genetic and nongenetic risk factors that may interact with one another. But studying these joint effects has been a challenge, largely because researchers lack the tools to identify and measure nongenetic risk factors, such as exposures to substances in the environment.
Nonetheless, many researchers believe that studying the interplay between genes and environmental exposures will be critical for understanding the biology of cancer and for preventing and treating the disease. This was the conclusion of a recent scientific meeting on gene-environment interactions hosted by the National Institutes of Health, to take one example.
“It’s tremendously important to understand the interplay between genes and environments,” said Dr. Muin Khoury, who directs the Office of Public Health Genomics at the Centers for Disease Control and Prevention (CDC). “If you can identify subgroups of individuals who are more susceptible to an environmental exposure than others, then you could develop targeted prevention strategies.”
Gene-environment interactions are worth studying for many reasons, added Dr. Duncan Thomas of the University of Southern California Norris Comprehensive Cancer Center, who has written about emerging approaches in the field. Investigating interactions can reveal important biological pathways in cancers and possibly point to new treatment strategies. “In principle, it should be possible to design a drug that exploits a gene-environment interaction,” said Dr. Thomas.
Better Evidence Needed
But at the moment many researchers are calling for more reliable data on interactions.
“Many gene-environment interactions in cancer have been proposed, but few interactions have rigorous support in the scientific literature,” said Dr. John Ioannidis, a professor of epidemiology at the Stanford School of Medicine. “This just means that there has been no systematic approach in the field until now.”
Large Studies, Reliable Data
Concerns about false positives in studies of gene-environment interactions are reminiscent of the early days of genome-wide association studies before the advent of very large, rigorous genome-wide association studies, noted Dr. Ioannidis.
“We need to move to an evaluation of gene-environment interaction on a massive scale,” he said. “And we need accurate reporting and careful grading of the evidence to overcome the ambiguity in the field.”
Dr. Khoury agreed. “To make progress, we need accurate information on longitudinal exposures and genomic data from large-scale epidemiology studies,” he said. “And that information is hard to come by.”
The concept of gene-environment interactions is not new, but there has been a resurgence of interest in them. One reason could be the wealth of new information on genetic risk factors. As several new studies suggest, this knowledge can be leveraged to gain insights into environmental risk factors and possible interactions.
“We believe these interactions could explain a lot of differences between groups in the risk of disease from environmental exposures,” said Dr. Deborah Winn, deputy director of NCI’s Division of Cancer Control and Population Sciences (DCCPS). “There may be people in specific populations who are susceptible. And by looking only at genes or at exposures, you are not going to explain as much as an interaction would.”
A traditional view of gene-environment interactions suggests that people with gene A may be susceptible to disease B after being exposed to something in the environment. But new statistical tools are needed to assess many “environmental” factors, including factors within the human body such as inflammation.
The development of these statistical tools is really in its infancy, noted Dr. Clarice Weinberg of the National Institute of Environmental Health Sciences, who stressed the importance of understanding the biological mechanisms that underlie interactions.
“If two risk factors are known to be related to cancer, then it becomes interesting to learn how they work jointly,” said Dr. Weinberg. “But, fundamentally, in order to advance public health we need to understand the biology and what these factors do.”
Searching for Interactions
The “true biology” of most cancers is going to be much more complicated than a single gene and a single environmental exposure, said Dr. Khoury, who is also affiliated with DCCPS, so “we need creative ways of looking at multiple gene-environment interactions.”
One strategy is illustrated by a new study of well-done meat consumption and the risk of colorectal polyps, which can be precursors for cancer. Meats cooked at high temperatures form chemicals that can damage DNA. Whether exposure to these chemicals poses a risk to people is unclear, but the new findings support the idea that some people, because of their genetic makeup, may be more at risk from these chemicals than others.
In the study, red meat appeared to be a stronger risk factor for colorectal polyps among people who carried certain genetic variants than among people who did not carry these variants, researchers at Vanderbilt-Ingram Cancer Center reported in the American Journal of Clinical Nutrition. But Dr. Wei Zheng, the study’s senior investigator, cautioned that the results need to be confirmed.
“This study is just not large enough to say clearly there is something going on and that people who carry more of these risk variants would benefit from changing their behavior any more or less than those with fewer risk variants,” commented Dr. Peter Kraft of the Harvard School of Public Health, who was not involved in the research.
Chemical mutagens in meats have been suspected of playing a role in colorectal cancer for years. What’s new about this study is that the authors developed a genetic “score” for each participant. The score was based on 16 genetic variants that have been linked to metabolizing heterocyclic amines (HCAs), which are chemical mutagens found in well-done meat.
To look for interactions, the study authors analyzed each participant’s genetic score along with information about the participant’s diet and history of colorectal polyps. “The study is a good example of the research that has been done over the past two decades on enzymes involved in metabolizing heterocyclic amines in red meat,” said Dr. Zheng.
Genes, Sodas, and Obesity
Another recent study used a genetic score to explore the effects of consuming sugar-sweetened beverages on the risk of obesity. For each participant, the researchers calculated a genetic predisposition score based on 32 DNA variants that researchers have linked to body-mass index.
Participants with certain genetic variants appeared to be more susceptible to the adverse effects of sugar-sweetened beverages on obesity, Dr. Qibin Qi of the Harvard School of Public Health and his colleagues reported in the New England Journal of Medicine.
We believe these interactions could explain a lot of differences between groups in the risk of disease from environmental exposures.
—Dr. Deborah Winn
This is “a clear example of gene-environment interaction,” Dr. Sonia Caprio of the Yale School of Medicine wrote in an accompanying editorial. She pointed out that the interaction was apparent only when multiple variants were used to calculate the genetic score.
“Putting multiple genetic variants together in these kinds of scores is both interesting and useful,” said Dr. Kraft. “It’s not a new idea to look for gene-environment interaction, but a decade ago we didn’t know what these gene variants were doing.”
Both studies emphasize the need for “better biomarkers that can assess environmental exposures,” added Dr. Thomas, who has studied chemical mutagens in meat but was not involved in either study. The tools don’t yet exist for accurately measuring most exposures either outside or inside the human body.
The meat study also shows just how complicated interactions can be. Heterocyclic amines form when meat is cooked at high temperatures, but the chemicals are initially inert. It is only after HCAs are metabolized in the body that they become potentially harmful to DNA. At the same time, the body has proteins that detoxify these chemicals, rendering them harmless.
As Dr. Zheng and his colleagues reported, genetic variants associated with the metabolization and detoxification of HCAs may help determine a person’s internal exposure to these chemicals. “The body has ways of getting rid of the effects of these toxic exposures, and some of us are better at doing that than others,” explained Dr. Weinberg.
Similarly, some people's bodies are better at repairing damaged DNA than others, and this may be a piece of the HCA puzzle. As another recent study of HCAs and colorectal cancer concluded, researchers may need to evaluate a number of genetic pathways to gain insights into interactions.
“Our findings show that we need to look at multiple pathways—such as those involved in cell signaling and DNA repair—in addition to the HCA metabolizing pathways,” said Dr. Rashmi Sinha of NCI’s Division of Cancer Epidemiology and Genetics (DCEG), who led that study.
Along with expanding the number of pathways tested, future studies should also try to include data on exposures that take place over many years, if not a lifetime, several researchers said.
Environmental exposures tend to change over time, and little is known about which exposures matter most. Is it the average exposure over 10 years, or the peak exposure, or just the last 2 years? For most diseases, no one knows. Many researchers believe, however, that certain times in life, such as fetal development or adolescence, may be critical for some exposures.
“It gets complicated,” said Dr. Kraft. “These issues are solvable, but it’s going to take very large studies to solve them. The more complicated the data, the larger the study needs to be.”
—Edward R. Winstead