The recent finding of seven genetic variants in African-American men that could contribute to the greater prevalence of prostate cancer in this group stems from a newly developed approach that can identify genetic risk for diseases that vary across ethnicities. But the scanning approach, called admixture mapping, is not a new idea. The 50-year-old concept of evaluating genetic differences within recently blended populations could be put into practice only in the past five years due to technological advances and research by a handful of HMS scientists, including David Reich, HMS assistant professor of genetics. Reich began working on the admixture mapping technique as a postdoctoral researcher at the Whitehead Institute.
“There were a few technological obstacles along the way,” Reich said, listing unavailability of genome information and inadequate statistical analyses as key roadblocks. The lack of genomic information meant that there were few genetic variants known to be useful markers for ethnicity. With the sequencing of the human genome completed in April 2003, it became easier to identify these variants, and Reich and his colleagues experimentally defined several thousand of them. Even with this resource, however, they still lacked a way to analyze the massive amount of data collected from thousands of people. With the help of statistician Nick Patterson at the Broad Institute, Reich wrote a statistical program to handle the data.
Armed with the tools necessary to test whether admixture mapping could locate disease genes, Reich—a population geneticist—has been able to search for these genes in prostate cancer and multiple sclerosis, disorders that are more prevalent in African Americans and European Americans, respectively. Last summer Reich and his colleagues published findings from a scan for prostate cancer genes in about 2,500 African-American men, identifying a small region constituting about a thousandth of the genome as containing genetic variants contributing to risk for the cancer.
With the admixture technique, “you don’t have to look in all the nooks and crannies of the genome,” Reich said. “You first use admixture mapping to identify a big chunk of the genome that you know must contain disease genes.” In African Americans, the mixing of European and African ancestry occurred within the last ten generations, so the chunks have not had much time to break up due to recombination and give approximate localization of the disease genes. After a successful admixture scan, “which gets you 99.9 percent of the way there, you can follow up by a concentrated, high-resolution scan to pinpoint what you’re interested in,” Reich said.
Based on their previous findings that genetic variants for prostate cancer risk were likely to be contained on a particular region of chromosome 8—as well as a parallel discovery by the Icelandic genetics company deCODE that independently found two specific variants in the region contributing to risk—Reich and his team sought to identify whether additional genetic variants in this region contributed to the increased risk for African Americans. The researchers found seven genetic variants in black men with prostate cancer, spread over three small regions, which all independently contribute to disease risk. The study was published online April 1 in Nature Genetics.
Reich’s study arose from a collaboration with epidemiologists Christopher Haiman and Brian Henderson from the University of Southern California, who are part of the Multi-ethnic Cohort Study of more than 215,000 people from Hawaii and the Los Angeles area. The HMS and USC researchers used genetic information obtained from blood samples collected from about 7,500 African-American, Japanese-American, Native-Hawaiian, Latino, and European-American men with and without prostate cancer.
“The seven genetic variations we found are all more common in African-American than in European-American men,” Reich said, “but they are present in all populations. If you could medically intervene to reduce the rate of prostate cancer in African Americans to the level that would be expected if no one in the population carried any of these risk factors (and there are some people in this category already), the rate of disease would drop by up to two thirds.”
The study does not indicate what is causing the cancer, however. “There are no known genes on that area,” Reich said. Because the seven risk factors are dispersed over three clusters, the mechanism underlying the cancer may be unconventional. “It is possible that these are long-range regulatory elements in these regions. Alternatively, the genetic changes may increase the probability that in a man’s prostate, there will be an increase in the number of copies he has of this part of his chromosome 8.” It is known that this increase in copy number is a common feature of aggressive prostate cancer tumors and thus could be a novel mechanism of risk.
