Making cancer a thing of the past

The self-described product of "a long line of science and math geeks," Dr. Christopher S. Carlson's roots in science run deep. Winning his first grade class' "mad scientist award" cemented early on his determination to become a real scientist, and after years of hard work, he is now one of the newest faculty members in the Public Health Sciences Division.

With a doctorate in genetics and a bachelor's degree in molecular biology, Carlson is poised to better understand the role genetics play in determining future risks. His lab is currently focused on analyzing correlations between genetic variations and common diseases, such as cancer, cardiovascular disease and neurological disease. Early in his career, he was instrumental in coming up with an algorithm that would allow researchers to more efficiently analyze genes by selecting single nucleotide polymorphisms (SNPs) from the genome based on their information content.

"Where there are something like 6 million common variants in the human genome, you don't need to genotype every one of those. As a matter of fact, if you do, you've wasted a lot of energy, because many of them are very strongly correlated with each other," said Carlson, who joined PHS last fall.

By analyzing more of the genome with fewer genotypes, researchers are able to maximize the bang for their buck. They're also able to understand larger populations and how to best prevent common diseases. Carlson uses the example of a 20-year study of adults who were 18-30 at the beginning of the study, but who are now 38-50. Most have not yet had heart attacks, but researchers are looking for early risk factors in this and similar studies.

"If you're going to develop a heart attack by the time you're 60, before then, we know your cholesterol is likely to be high. We know your blood pressure will be high. So these are intermediate risk factors that develop over a lifetime. What we're trying to do is predict those intermediate risk factors from the SNP genotypes," Carlson said.

Predictive protein

To better predict likelihood of disease, the researchers are getting some help from a protein known as C-reactive protein (CRP), which is associated with inflammation. Traditionally, clinicians who wanted to know an individual's risk of heart disease would assess gender, cholesterol, hypertension status and whether the individual was a smoker or a diabetic. But some people with those risk factors do not develop heart disease, while others without the risk factors fall victim. Although CRP is correlated with all of those risk factors, it provides additional risk information, and can provide greater sensitivity in predicting whether an individual will develop cardiovascular disease.

As a postdoc, Carlson worked on CRP in Dr. Debbie Nickerson's lab at the University of Washington. During his time there, he was looking for SNPs, or tiny variations in DNA that can be used to track inheritance in families. By looking at differences in alleles, the variant forms of genes responsible for things like eye and hair color, researchers can zero in on how groups of people are related. In sequencing 47 people — half African American and half European American — the lab found 18 common variants, which Carlson distilled into seven tag SNPs that best captured the common variation. He found that three of the seven actually predict CRP levels.

"When we found three polymorphisms that correlated with CRP, at each of the polymorphisms, there's an allele that correlates with higher CRP levels, and an allele that correlates with lower CRP levels. For all three of these polymorphisms, the allele that correlated with higher levels was found at higher frequency in African Americans," Carlson said. "African Americans are known to have higher average levels of CRP than European Americans. Remarkably, after adjusting for BMI and genotype at these three SNPs, there was no residual effect for ethnicity."

Because the differences between ethnicities in CRP was explained either through body-mass index (BMI) or genetic makeup of the three SNPs, differences in allele frequencies at functional SNPs play a significant role in ethnic differences in disease risk. "The traditional assumption has been that there is a socio-economic variable or some other environmental parameter that we're not measuring correctly, and that this unmeasured variable is responsible for the difference in disease risk between ethnicities. In this case, there is a genetic component; what matters is your genotype at the functional SNPs, not your ethnicity, per se," Carlson said, adding that more studies need to focus on non-European populations.

Genetic-ancestry study

Carlson is interested in studying genetic ancestry and how our ancestors determine our own risks for disease. For example, diabetes is more prevalent in African Americans than Europeans. Geneticists believe that some of this risk is likely inherited for evolutionary reasons. Mitochondria in cells of Africans are typically more efficient than mitochondria in Europeans and waste less energy as heat. "Efficient use of calories is great news when you're just trying to survive in a tropical climate," Carlson said. "It's bad news when you're in an environment where calories are free."

So if your ancestors lived in an area prone to famine, they were more likely to have a "thrifty" genotype that predisposed them to stow away calories so they could survive the famine cycle. "Those would have been advantageous in a frequent famine environment; in the modern world, they're not. Genotypes that better survive famine would have higher risks of obesity and Type II diabetes on a modern diet," Carlson said.

Populations that survived epidemics like smallpox could also prove to be very interesting to Carlson. Many of our anti-viral immune responses involve detection of dysregulated cells, so immune systems are set up to destroy unregulated cells, or the cells will self-destruct. "Those same surveillance mechanisms work on cancer," Carlson said. "You don't want cells to be dividing in an unregulated fashion, whether it's a cancer or a virus that has commandeered the cell. Either way, you want to take them out." Carlson hypothesizes that genotypes that conferred resistance to viruses could well have a relationship to cancer resistance.

Public-health repercussions

Translating ancestral background into cancer prevention could have broad repercussions in determining the best course of action for various populations. "From a public-health standpoint, if we understand the pathways by which the SNPs that change things modestly contribute to your disease risk, this gives us something where maybe we can start interfering with those pathways on a global scale," Carlson said.

One of the reasons he is most excited about doing research at the Center is the access to broad populations. Studying large groups of people is invaluable to reaching conclusions about how to keep cancer at bay in targeted populations. There may be drugs and prevention measures that simply work better for certain genotypes, but those are still some time away.

"I'm not confident that genetic diagnostics will give us adequate information about whether an individual patient will develop disease," Carlson said. "But I am confident that it will give us insights into new ways to better tackle the diseases, and hopefully into how to treat an individual patient."

Carlson's work is funded by several grants, including support from two UW grants to other principal investigators. He is also supported by the Cancer Center Support Grant (CCSG) and recently received a grant from the National Cancer Institute.