Genetic variation among races, long a political hot potato, has also been a scientific puzzle.
Although researchers have cataloged different frequencies of inherited DNA among racial groups, and physicians have found that some groups are disproportionately susceptible to certain diseases, it’s not clear how or even whether the two are linked. Do subtle differences in DNA between races really matter, medically speaking?
Earlier this week, scientists described results from a new approach that may help answer that question: measuring gene expression levels among Caucasians and Asians.
Because gene expression helps determine how a cell behaves, it can be more instructive than variations in inherited DNA. The researchers examined expression levels of more than 4000 genes in 142 banked cell lines drawn from individuals of European descent in Utah, and cohorts from Beijing and Tokyo. They found that 25% of the genes had expression patterns with statistically significant, although often small, differences depending on whether they came from a Caucasian or an Asian sample. Thirty-f ive genes had expression levels that differed, on average, as much as twofold. Still, “how that translates into traits of clinical interest is still a big question mark,” says Neil Risch, a human geneticist at the University of California, San Francisco.
Although that critical bridge remains to be built, scientists say the expression patterns are intriguing. Indeed, geneticist Vivian Cheung of the University of Pennsylvania, who led the research team with her colleague Richard Spielman, was initially so taken aback by the number of genes whose expression varied that she suspected a technical glitch. “The 25% definitely shocked me,” says Cheung, who also works at the Children’s Hospital of Philadelphia.
But when she and her colleagues repeated the study on samples from 24 Chinese residents in Los Angeles, the results were virtually identical. All but one of the 35 genes with big variations in expression registered similar levels in the HapMap Asian samples and the Los Angeles cohort, they report online this week in Nature Genetics.
“This lends support to the idea that there are genetically determined characteristics that tend to be clustered in different ethnic groups,” says Phyllis August, a nephrologist at Weill Medical College of Cornell University in New York City, who has studied variation between blacks and whites in a gene involved in hypertension.
“To deny that is really denying a lot of very obvious biological truths.” Researchers are careful to say that although mean expression between Asians and Caucasians differed in more than 1000 genes studied, the expression difference between individuals from each group was often not impressive. “These averages are not absolutes,” says Stephen Wooding, a population geneticist at the University of Texas Southwestern Medical Center in Dallas. He compares the variation in gene expression to height in men and women; although men on average are taller, plenty of individual women are taller than individual men.
To analyze expression levels, Cheung and her colleagues began with samples collected for the International HapMap Project, which aimed to catalog genetic variation to help identify disease genes.
They used microarray technology to measure gene expression in several thousand genes at once and found measurable expression in 4197 genes.
Then, they compared mean expression levels in the three different sets of samples.
At first, the researchers separated the Chinese and Japanese samples but then lumped them together after finding that only 27 genes registered different mean expression levels between the two. The different expression levels seemed to correspond to patterns of
Inherited variation in single-nucleotide polymorphisms (SNPs)—for example, if one DNA stretch with a particular SNP was rare in a higher percentage of Asians than Caucasians, average gene expression in the first group might be lower. It’s still not clear whether the SNPs themselves might be regulating gene expression, or whether they travel together with other DNA that’s the regulator.
The question now is whether and how these expression differences affect health.
One gene, called UGT2B17, is deleted more often in Asians than Caucasians and had a mean expression level that was 22 times greater in Caucasians than Asians, the most dramatic variation seen. “That one really stuck out,” says Wooding, who notes that this gene is involved in steroid metabolism and, possibly, drug metabolism as well.
Spielman agrees that genes such as UGT2B17 and others that showed up in the list of 35 should be looked at individually to determine what the expression differences might mean. Next up for his group: examining gene expression in other ethnicities, including Africans, to see what patterns materialize. –JENNIFER COUZIN