Darwinian dogma states that in the marathon race of evolution, the genotype that replicates the fastest, wins. But now scientists at the California Institute of Technology say that's true, but when you factor in another basic process of evolution, that of mutations, it's often the tortoise that defeats the hare.
It turns out that mutations, the random changes that can take place in a gene, are the wild cards in the great race. The researchers found that at high mutation rates, genotypes with a slower replication rate can displace faster replicators if the former has a higher "robustness"—or fitness—against mutations; that is, if a mutation is, on average, less harmful to the slower replicator than to the faster one. The research, to appear in the July 19th issue of the journal Nature, was conducted by several investigators, including Claus Wilke, a postdoctoral scholar, Chris Adami, who holds joint appointments at Caltech and the Jet Propulsion Lab, Jia Lan Wang, an undergraduate student, Charles Ofria, a former Caltech graduate student now at Michigan State University; and Richard Lenski, a professor at Michigan State.
In a takeoff of a common Darwinian phrase, they coin their work "survival of the flattest" rather than the survival of the fittest. The idea is this: If a group of similar genotypes with a faster replication rate occupies a "high and narrow peak" in the landscape of evolutionary fitness, while a different group of genotypes that replicates more slowly occupies a lower and flatter, or broader, peak, then, when mutation rates are high, the broadness of the lower peak can offset the height of the higher peak. That means the slower replicator wins. " In a way, organisms can trade replication speed for robustness against mutations and vice versa," says Wilke. "Ultimately, the organisms with the most advantageous combination of both will win."
Discerning such evolutionary nuances, though, is no easy task. To test an evolutionary theory requires generations and generations of an organism to pass. To make matters worse, the simplest living system, namely that which has been a precursor to all living systems on Earth, has been replaced by much more complicated systems over the last four billion years.
Wilke and his collaborators found the solution in the growing power of computers by constructing, via a software program, an artificial living system that behaves in remarkably lifelike ways. Such digital creatures evolve in the same way biological life forms do; they live in, and adapt to, a virtual world created for them inside a computer. Doing so offers an opportunity to test generalizations about living systems that may extend beyond the organic life that biologists usually study. Though this research did not involve actual living organisms, one of the authors, Richard Lenski, is a leading expert on the evolution of Escherichia Coli bacteria. Lenski believes that digital organisms are sufficiently realistic to yield biological insights, and he continues his research on both E. coli and digital organisms.
In their digital world, the organisms are self-replicating computer programs that compete with one another for CPU (central processing units) cycles, which are their limiting resource. Digital organisms have genomes in the form of a series of instructions, and phenotypes that are obtained by execution of their genomic program. The creatures physically inhabit a reserved space in the computer's memory—an "artificial Petri dish"—and they must copy their own genomes. Moreover, their evolution does not proceed toward a target specified in advance, but rather proceeds in an open-ended manner to produce phenotypes that are more successful in a particular environment.
Digital creatures lend themselves to evolutionary experiments because their environment can be readily manipulated to examine the importance of various selective pressures. In this study, though, the only environmental factor varied was the mutation rate. Whereas in nature, mutations are random changes that can take place in DNA, a digital organism's mutations occur in the random changes of its particular computer program. A command may be switched, for example, or a sequence of instructions copied twice.
For this study, the scientists derived 40 pairs of digital organisms that were derived from 40 different ancestors in identical selective environments. The only difference was that one of each pair was subjected to a four-fold higher mutation rate. In 12 cases out of the 40, the dominant genotype that evolved at the lower mutation rate replicated at a pace that was 1.5-fold faster than its counterpart at the higher mutation rate.
Next, the scientists allowed each of these 12 disparate pairs to compete across a range of mutation rates. In each case, as the mutation rate was increased, the outcome of competition switched to favor the genotype that had the lower replication rate. The researchers believe that these slower genotypes, although they occupied a lower fitness peak and were located in flatter regions of the fitness surface, were, as a result, more robust with respect to mutations.
The digital organisms have the advantage that many generations can be studied in a brief period of time. But the researchers believe a colony of asexual bacteria, subjected to the same stresses as the digital organisms, would probably face similar consequences.
The concept of "survival of the flattest" seems to imply, the authors say, that, at least for populations subject to a high mutation rate, selection acts upon a group of mutants rather than the individual. Thus, under such circumstances, genotypes that unselfishly produce mutant genotypes of high fitness are selected for, and supported in turn, by other mutants in that group. The study therefore reveals that "selfish genes," while being the successful strategy at low mutation rates, may be outcompeted by unselfish ones when the mutation rate is high.