How to tell the age of the extinct ancestor common to two living species? Take the same gene from both and count the differences in the DNA. Then divide it by the rate at which DNA mutates, and presto! Biologists love this molecular clock and they've applied it for decades. But new research shows it may not be keeping good time.
In 1965, biochemists stunned classical biologists by showing that mutations in genes accumulate at a constant rate. Hence, the number of mutations between two species could be used to tell when species diverged, even without fossil evidence. Since then, scientists have used molecular clocks in various genes to trace evolution in species from HIV to birds to whales. And although exceptions to the rule have surfaced, most biologists remain loyal to their favorite timepiece.
But in the 25 September issue of the Proceedings of the National Academy of Sciences, geneticist Francisco Rodriguez-Trelles and his colleagues at the University of California, Irvine, show that the molecular clock could be ripe for the pawnshop. From the GenBank database, they downloaded the sequences of three well-known genes, called Gpdh, Sod, and Xdh, for 78 species, from pine trees to people. They used the data to make an evolutionary tree, which, for calibration purposes, included some key branches already dated by paleontologists.
But when the scientists started counting the number of mutations in each tree branch, they found vastly different mutation rates, even for closely related species. For example, in Drosophila obscura fruit flies, the Sod clock ticks 10 times faster than in its cousin Drosophila willistoni, while the Xdh clock keeps perfect time. Conversely, the Gpdh clock in mammals runs about 10 times faster than the one in fruit flies. Molecular clocks in general are much more "erratic" than previously thought, and practically useless to keep accurate evolutionary time, the researchers conclude. They attribute this to the vagaries of natural selection, which may at times constrain specific genetic mutations in certain lineages.
Evolutionary biologists are unhappy to hear that such a prized correlation may be flawed. David Mindell, of the University of Michigan, Ann Arbor, says it's "bad news that estimates of dates must be viewed as highly error prone." But, he adds, ultimately it's good news that, by disproving the molecular clock theory, molecular evolution in a wide variety of organisms may become better understood.