Feature: Replaying Lifes Tapes
In a novel experiment, Michigan State University researchers test the repeatability of evolution. In a series of bacteria-filled flasks, scientists at Michigan State University are searching for answers to one of the most difficult questions in evolutionary biology.
If the history of life could be replayed by somehow reviving ancient organisms and putting them into the same environments they experienced for millions of years, would their descendants be any different from those that actually evolved? Some biologists, who regard adaptation as the primary force in evolution, think such organisms would be largely the same, as a result of their similar environments. But others, who contend that the history of life was uniquely modified by chance fluctuations in a population's genes and other random events, believe that evolution is irreproducible. The debate has sharply divided evolutionary biologists.
One of the most prominent proponents of the latter view is Stephen Jay Gould, a paleontologist at Harvard University, who wrote in his book Wonderful Life: The Burgess Shale and the Nature of History (W. W. Norton & Company, 1989) that 'replaying life's tape' over and over again would consistently 'lead evolution down a pathway radically different from the road actually taken.'
'Each step proceeds for cause,' Gould added, 'but no finale can be specified at the start, and none would ever occur a second time in the same way, because any pathway proceeds through thousands of improbable stages. Alter any early event, ever so slightly and without apparent importance at the time, and evolution cascades into a radically different channel.'
A persuasive argument, but is it actually the way evolution proceeded? While no one can repeat the entire history of life, biologists at Michigan State have done the next best thing. They have provided new fuel for the debate by replaying a portion of evolution in their laboratory.
AN UNUSUALLY RAPID SYSTEM
In a series of flasks, the scientists have simulated a segment of what Mr. Gould calls 'life's tape' using the common bacterium E. coli. Such an organism, which takes only hours to produce a new generation, provides an unusually rapid evolutionary system, enabling scientists to view evolution over many generations in a short time. The continuing experiments have produced more than 15,000 generations of microbes in six years. That's not all. Because the bacteria can be cloned -- unlike fruit flies and other organisms commonly used in genetic experiments -- the researchers also are able to use the same ancestor for many different experiments. 'When we start these replicate populations to ask how reproducible evolution is, we start each population from, in essence, a single cell,' says Richard E. Lenski, professor of microbiology and zoology. 'So there is no genetic variation at time zero, and all of the variation has to accumulate de novo, by mutation.'
Freezing the ancestral bacteria provides another important bit of information. 'Resurrecting the bacteria from the dead,' as Lenski puts it, has allowed the researchers to compare the ancestors with their descendants to determine how evolution has changed the descendants.
In their first experiments, Lenski and Michael Travisano sought to examine the relative impact of chance and adaptation on evolution. They started with 12 genetically identical populations of bacteria placed into the same new environment, one that forced the bacteria to change in order to survive on a new type of food. The degree of those changes provided the researchers with a glimpse of the relative importance in evolution of random events (which tend to make the populations differ from one another) and adaptation (which tends to make the bacteria evolve to similar forms because of their identical environments).
What they found, after 10,000 generations, was that descendants were able to reproduce 50 per cent faster, on average, than their ancestors in the new environment -- indicating that they were far better adapted. But genetic fitness -- that is, the ability of the bacteria to reproduce in their new environment -- varied slightly among the 12 lines of descendants. The fact that there was any variation at all is significant. One would expect genetically identical bacteria, placed in an identical new environment, to have the same genetic fitness if adaptation were the only force in evolution. But because they don't, random events also must be involved. How big a role adaptation and random events play in evolution can be determined by the differences in genetic fitness among the 12 descendant populations -- which, Lenski says, 'averaged about 10 per cent as much as the average difference between them and the ancestor. So they all increased dramatically to a similar extent, but there were still subtle differences in genetic fitness that crept in among the populations.'
Those differences suggested to the scientists that the chance events could alter the course of evolution. A second, more complex experiment reinforced that suspicion. In that experiment, Travisano, Judith A. Mongold, and Lenski, all of Michigan State, working with Albert F. Bennett at the University of California at Irvine, sought to determine the repeatability of evolution by examining the relative contributions of adaptation, chance events, and history to evolutionary change.
INHERITED CONSTITUTION
Evolutionary biologists have long wondered whether an organism's previous evolution -- what they refer to as its 'history' -- alters its subsequent evolution. some scientists, such as Harvard's Gould, believe that an organism's range of possible adaptations is limited by its inherited constitution, which constrains the number of genetic combinations that can develop from its ancestor's genes.
The results of the experiment, which were published this year in Science suggest that evolution in not repeatable, although initially this wasn't apparent to the scientists. What they found was that an organism's adaptation to its environment quickly overwhelms random events and its genetic past, but that chance can play a dominant role in those evolutionary aspects -- such as the size of a bacterial cell -- that have little impact on its ability to survive or reproduce. Given enough time, the genetic variation produced by chance inevitably leads to a large number of changes in the organism, making it impossible exactly to repeat the process of evolution. 'Natural selection is such a powerful homogenizing force that you have to watch it for a very long time, and very carefully, to become convinced that it wouldn't be reproducible,' says Lenski. 'The differences are pretty subtle, and it's only when you get to the conclusion and those little differences matter that the whole ending is irreproducible.'
'Over time,' he adds, 'enough of these chance events creep in so that the cumulative effect of a lot of small and seemingly inconsequential differences means that if you look at the evolutionary pathway of a population, evolution becomes quite irreproducible over a very long period of time.' Gould says he hasn't yet read the Science paper, but adds, 'It sounds very interesting.'
TESTING ANOTHER ASPECT
While Lenski concedes that a more lifelike setting might have produced different results, he and his colleagues aren't finished with their experiments. They hope to test one other aspect of evolution: whether major improvements can occur through a random shuffling of genes. After allowing the bacteria to reproduce for more than 15,000 generations in their new environments, they found that improvements in genetic fitness in the colonies -- measured by their rate of reproduction -- basically stopped.