|Prime time for sequencing primate genomes|
By Adam Marcus
March 7, 2003
Of the hundreds of genomes sequenced so far only one, ours, belongs to a member of the primate family. The sequencing of the chimpanzee genome, begun this winter, will change that. But most of the species in the sequencing pipeline are not closely related to humans.
That's no accident. Many scientists have thought the best way to make sense of the human genome is to identify DNA that is conserved between humans and species like the mouse, the pufferfish and the worm.
Some also thought that because primates are so genetically similar to peoplechimps and humans share nearly 99 percent of the same DNAside-by-side comparisons of sequences would be too difficult.
Now, it seems that the field of genomics may be ripe for a shift in focus. New methods of studying highly similar genomes should allow researchers to compare primate genomes to learn what makes the human genome uniqueincluding the DNA that turns genes on and off.
One such method, known as "phylogenetic shadowing," is a kind of genomic triangulation that can pick out genes and DNA that regulates genes among primate genomeseven when the DNA is nearly identical.
Researchers led by Edward Rubin, of the U.S. Department of Energy's Joint Genome Institute, in Walnut Creek, California, recently used the strategy to find similarities between human DNA and that of 17 close primate relatives, including chimps, orangutans and baboons.
In principle, as few as four or six other primates should be enough to make meaningful comparisons, say the researchers, whose findings appear in Science.
Rubin's team also showed that following the evolutionary trail of a specific gene could reveal regulatory instructions that surround it. The gene they chose, apolipoprotein (a), occurs only in Old World monkeys and humans. It produces a blood protein that in high levels is a marker for heart disease.
If the goal is to understand the human genome, says Rubin, "trying to sequence a group of primates and not just one will provide us a lot of information about what uniquely builds a human."
"I'd like to get a data set of a bunch of primate genomes, lay it on the human genome and see what's conserved," he says. This would reveal non-gene regions that may regulate genes. "That which is important in one species' genome tends to be conserved in another."
Roger Bumgarner, a geneticist at the University of Washington, in Seattle, says the approach has "real advantages" over the conventional method of comparing distantly related genomes. When looking at the human and mouse genome, for example, it's often difficult to distinguish useful information from false leads, Bumgarner says.
"This method seems to reduce the number of false positives pretty significantly," he adds.
Evan Eichler, a human geneticist at Case Western Reserve University in Cleveland, says that Rubin's method should help researchers tackle one of the biggest question of next-generation genomics: what are the regulatory elements that control the expression of genes and how do they work.
By comparing humans with near relatives, "you might find things that you might have missed with more distant comparisons," Eichler says.
Here's an example. Scientists long believed that occasional one-letter differences in the genetic code (known as single nucleotide polymorphisms) account for most of the genetic difference between humans and non-human primates. This variation, the model held, was the key to why genes are expressed differently in humans and non-human primates.
Now, by comparing primate DNA, researchers have found that a significant amount of the differences between humans and non-human primates can be accounted for by small differences in DNA that is either present (inserted) or absent (deleted) within and around genes. The study, led by Kelly Frazer of Perlegen Sciences in Mountain View, California, appears in Genome Research.
Why and where these changes occur isn't clear, Frazer says, though she believes the process is random.
Frazer calls her study "the first direct evidence" that the human genome contains substantial stretches of DNA, as long as ten thousand base pairs, that are unique to the species. "It's probably not so much that there are genes that are different [in the spans] but they might be involved in differential expression" of genes, she says.
However, Frazer's group did find one gene, for a cell surface protein used to diagnose cancer in people that may have been deleted in at least one primate. The gene is in chimps, orangutans and macaques but not in woolly monkeys. If the gene is truly absent in that creature, she says, it would be proof that deleted DNA can result in functional changes.
Richard Gibbs, a geneticist at Baylor College of Medicine in Houston, Texas, says there's plenty of enthusiasm in the field for sequencing primates"but not a full set like Rubin would have."
Gibbs, co-author of a commentary on the Science paper, says the work "gives us a new paradigm to think about, but we're not going to drop everything else and just do primates."
Indeed, the National Human Genome Research Institute has so far agreed to fund genome sequencing for only two primates, the chimpanzee and the rhesus macaque.
Rubin, for his part, believes he's hit on the right approach. "I've been battling people trying to convince them that we should be sequencing more primates," he says.
. . .