Professor Richard Lenski of Michigan State University started something big. He had the grand vision, patience, and perseverance necessary to design and conduct a prospective interventional study of evolution among bacteria. As we will see, this study was brilliantly designed to produce very-high-confidence results. Lenski’s study remains one of the best, if not the best, high-confidence scientific study of evolution that has ever been conducted.
It started with twelve identical populations of bacteria called Escherichia coli (E. coli). Lenski called these bacteria the “founding fathers,” because they were pioneers, set apart in a very controlled environment in which it would be easy to observe changes over time. Lenski split the bacteria into twelve populations to see if they would evolve in similar or divergent directions. The controlled environment consisted of a warm bath of a solution called Davis Minimal Broth with a small amount of added glucose as the source of energy. The glucose and other nutrients were intentionally limited to establish a competition for the bacteria. This limitation of nutrients was a very important part of the experiment. The bacteria that were faster to grow, or the bacteria that could make use of nutrients that others could not, would dominate. This is natural selection—survival of the fittest. Although the bacteria were free to partake of the limited nutrients, their rapid reproduction and growth quickly depleted the available nutrients within a twenty-four-hour period. At that point, after six or seven generations of E. coli had been produced, reproduction and growth came to a halt. To keep the experiment going, a small portion of each flask (containing a small fraction of the bacteria from the previous day’s growth) was withdrawn and added to a new flask with a fresh food supply; this happened every twenty-four hours. The experiment is now nearing its 10,000th day (almost thirty years!) of continuous operation, with a total of 67,000 generations of E. coli studied. On every seventy-fifth day, a portion of the bacteria from each flask is frozen, which is another very important part of the experimental design. These bacteria can be frozen in time and can be revived at will in order to repeat or replay a portion of the experiment. Perhaps you’ve heard that repeatability is an important part of high-confidence science.
We learned above that each nucleotide of DNA that is copied has about a 1 in 100 million (108) chance of a point mutation. Thus, if 100 million organisms are produced, chances are that each nucleotide in the genome will have experienced a point mutation in at least one of the organisms. The total number of E. coli organisms that have lived under Lenski’s experiment is roughly 6X1013 (60 trillion). Thus, it is almost certain that each and every one of the approximately 5 million base pairs of DNA in the E. coli has been changed in at least one organism during the experiment. This can be viewed as an exhaustive test to find any potential benefit of any single point mutation.
The results? It is very clear that changes have occurred. According to our generalized definition of evolution—changes in the properties of organisms that occur over more than one lifetime—we can conclude with very high confidence that generalized evolution has occurred. The bacteria adapted to compete in this particular environment. The size of the bacteria increased over time, their growth rate on glucose increased, and their speed of exiting a dormant state and returning to a growth state when introduced to a new flask increased. However, these changes quickly reached new plateaus, which demonstrates the limits to how far E. coli can be stretched by this environment. Numerous research papers have been written about these changes.42 Although the changes that occurred tended to be beneficial in this particular environment, this does not imply that the changes would be beneficial in a more natural setting. Several of the changes are likely to be detrimental to bacteria if they are placed back in a natural environment. For example, four of the twelve strains developed defects in their DNA repair mechanisms, and, after only two thousand generations, all twelve strains lost the ability to manufacture D-ribose, a component of RNA.
Our focus will be on the most significant change that occurred, because we are interested in knowing how far generalized evolution can go. After about 33,000 generations, one of the daily flasks of bacteria was surprisingly cloudy. This could be either a result of contamination or some important change in the E. coli. Lenski’s team was able to rule out the presence of contaminants so that they could conclude that the E. coli had changed in some significant way. The cloudy appearance of the flask stemmed from a greatly increased population of E. coli. Given the limited availability of glucose, which previously constrained population sizes, the E. coli must have found another source of energy in the flask. The limited nutrient bath contained one likely new source of energy: citrate. Citrate was abundant in the bath, but it was not intended to be a nutrient—it was included as a “chelating agent,” which is a way of providing a protective packaging for metal ions in solution. By extracting some of these E. coli and placing them in a flask with only citrate as a nutrient, the study authors were able to prove that the E. coli had indeed evolved the ability to metabolize citrate. It is well known among biologists that E. coli cannot metabolize citrate when oxygen is present,45 which is a feature that helps to define them as a species.46
In their first report of novel citrate metabolism in E. coli,47 the authors did not know the specific genetic changes that were responsible but determined that the changes must have involved at least two mutational steps. They also realized that the acquisition of this capability was extremely rare—more than 10 trillion (1013) E. coli in the earlier generations had failed to develop the ability to metabolize citrate.
The discovery of a newly evolved metabolic path sent ripples through the scientific community. Although the first research paper described the results with sober judgment (one of our six criteria of high-confidence science), a summary by Dawkins greatly amplified the findings (emphasis added; “they” refers to creationists):
Not only does it show evolution in action; not only does it show new information entering genomes without the intervention of a designer, which is something they have all been told to deny is possible; not only does it demonstrate the power of natural selection to put together combinations of new genes that, by the naive calculations so beloved of creationists, should be tantamount to impossible; it also undermines their central dogma of “irreducible complexity.”48,49
Dawkins made these bold assertions before the actual genetic changes were determined. His claims that the changes included “new information entering genomes” and “combinations of new genes” were actually hopeful predictions at this point, not established observations.
Over the next seven years the exact genetic mechanisms behind the change were published in a series of two papers.50,51 First, a little background is in order. Normal E. coli are able to metabolize citrate, but only if oxygen is not present. Specifically, a single step in the process of metabolizing citrate—that of transporting the citrate molecule into the cell—does not happen in the presence of oxygen. This is because the gene that produces the citrate transporter has a promoter region that is inactivated by oxygen. In the absence of oxygen, this gene is turned on, and the cell can actively transport citrate across the cell membrane.
The first of two mutations occurred at approximately generation 31,500 of Lenski’s experiment: a gene for citrate transportation was copied to a new location. As mentioned earlier, the original copy was preceded by a promoter region that was disabled by oxygen. The new copy was preceded by a promoter region that was active in the presence of oxygen. Thus, the citrate transporter could now be produced in the presence of oxygen. This mutation on its own allowed the cell a very limited use of citrate in the presence of oxygen, because for every citrate molecule to be used as fuel, one succinate molecule had to be ejected from the cell as a trade.
The second mutation occurred at approximately generation 33,000. This mutation occurred in the promoter region of a second transporter—one that brings succinate into the cell. The mutation resulted in an eleven-fold increase in expression of this second transporter. The two mutations therefore work synergistically to bring citrate into the cell, because the cell can now bring in succinate and use that as payment to bring citrate into the cell. This process can provide energy, as long as there is citrate and succinate in the medium. The article by Erik Quandt and colleagues showed that these two mutations were sufficient on their own to allow the metabolism of citrate in the presence of oxygen.
To summarize what happened, two relatively minor changes in E. coli DNA were sufficient to allow for the metabolism of citrate: the first change involved duplicating an existing gene, and the second involved increased activity of a second existing gene. The first mutation offered a slight benefit, which was important in facilitating the acquisition of the second mutation. These changes occurred over 33,000 generations and approximately 10 trillion, or 1013, total organisms. As it turns out, a previously reported case of E. coli that could metabolize citrate in the presence of oxygen had been reported in 1982.52 Interestingly, this case also found that the change resulted from two mutations, although they were different than the two mutations in the Lenski experiment. In light of these findings, we can see that Dawkins’s bold claims, which he made before before the actual mechanism had been determined (“new information entering genomes” and “combinations of new genes”) were a gross and irresponsible exaggeration of reality.He has again clearly exposed himself as a salesman, not a scientist.
Although the ability to metabolize citrate in the presence of oxygen is clearly beneficial for the particular environment of the Lenski experiment, this does not imply that these two mutations are beneficial in a more natural setting. Both mutations represent a loss of control of gene expression, which can lead to the wasteful production of protein transporters when they are not needed. The six criteria for high-confidence science, Lenski’s study is trying to answer the question: What evolutionary changes occur in E. coli over time, given a controlled and constrained environment? The table below summarizes the argument that Lenski’s study is based on high-confidence science.
Stadler, Rob. The Scientific Approach to Evolution: What They Didn’t Teach You in Biology (pp. 101-102). (Function). Kindle Edition.
