Professor Lenski’s Amazing Citrate-Eating E. coli

You’ve heard of Richard Lenski’s long-term E. coli experiment before, I assume. It’s widely cited as a counterexample to a subset of creationist claims, and is particularly famous for Andrew Schlafly’s hillarious attempt to challenge Lenski’s findings in 2008. RationalWiki summarises what you should already know:

On June 9, 2008, the New Scientist published an article describing preliminary results of a long-running experiment started by Lenski [ED: link]. Lenski and his team had taken a single strain of the bacterium E. coli, separated its descendants into twelve populations, and proceeded to observe their mutations over the course of twenty years (a process discussed on Lenski’s website). The E. coli were fed a measured amount of glucose every day. At one point, one of the populations exploded far beyond the parameters of the experiment. Lenski eventually discovered that this population had evolved the ability to “eat” citrate, an organic molecule which was part of the solution the E. coli lived in, but which E. coli cannot normally digest. Thus, evolution had been visibly observed, with an exquisite amount of evidence establishing the timeline along the way. Not only that, but the experiment was repeatable; Lenski started new experiments with the frozen “archives” of the population which exploded, and found that beyond a certain point, that particular population of E. coli were highly likely to evolve the ability to digest citrate. The paper also highlighted the role of historical contingency in evolution and the role of potentiating mutations.

There is an additional, less commonly known but still important fact to note: your average E. coli can already metabolise citrate, just not in the presence of oxygen. Indeed, not doing so is a defining characteristic of the species* – does this experiment then qualify as the creation of a new “kind” of organism, I wonder?

Anyway, that’s what we already knew. While the appearance of citrate metabolism was well-documented, with huge quantities of data to wade through only now (with a new paper in Nature) do we have an insight into the changes at the genetic level that allowed the phenotypic changes be observed. Perhaps in a desire to become the next Schlafly, Brian Thomas writes Bacterial ‘Evolution’ Is Actually Design in Action.

There is a common counter to the experimental results that claims that because the experiment happened in a lab under human supervision, therefore design was involved and evolution cannot be credited for what is observed – this is silly. Slightly better, though not by much, is the claim (whether explicit or implicit) that God/The Most Exalted Designer reached down and screwed with Lenski’s experiment, inducing the necessary mutations and contaminating the results. Despite his choice of title, Thomas is not actually advocating either of these positions. Instead, in a manner reminiscent of Friday’s article, he argues that the bacteria were designed to design themselves. Remind you of anything?

He opens:

The evolutionary community has been buzzing over bacteria’s new ability to obtain citrate, a carbon-containing chemical, from their environment and use it as a food source. Some say this confirms evolution in action but what if the bacteria were designed to modify themselves? That might disappoint evolution enthusiasts.

There are a number of minor problems here – why did Brian bother to specify that citrate is ‘carbon-containing’? And how dare he imply that our nefarious plans for world domination are just some kind of a hobby? At any rate, showing that this is all a result of design is a pretty tall order. Skipping ahead:

To verify that this new trait represents the kind of evolution that can change a microbe into a man, researchers needed to describe exactly what went on behind the scenes.

As you can see, one of Brian’s tactics here is to differentiate between different types of evolution, to make it easier to dismiss the findings. In reality, there is no one “kind of evolution that can change a microbe into a man” – all processes, from horizontal gene transfer to speciation, will have contributed at some point.

Prior to this new trait, none of the other observed changes that Lenski and his team had tracked showed that the bacteria evolved into a different basic kind—they were all still E. coli. And so far, none of those changes specified any new functional coding, which would occurthe [sic – something seems to be missing here] same as if chance-based phenomena could write a computer program. Did the strain of E. coli that acquired the new ability to import citrate—called Cit⁺—construct new functional, biochemical machinery by chance-base mutations?

At face value this paragraph seems to imply that the new changes to E. coli have actually made a new ‘kind,’ which I doubt was his intention. Before we get any further here we need to take a trip into Behe land:

In 2010, biochemist Michael Behe reviewed 12 new phenotypes, which are outward expressions of genetic coding, that Lenski’s E. coli displayed from 1994 to 2008. Behe categorized the known genetics producing each new bacterial phenotypes as either losing, shuffling, or gaining what he called “functional coded elements,” which include genes and gene promoters. All the known changes in the bacteria were either a loss or reorganization of pre-existing functional coded elements. None of the new phenotypes came from a gain of functional coded elements, and yet this is what molecules-to-man evolution requires.

At that time, the mechanism underlying the citrate-eating phenotype was unknown. Behe wrote, “If the [Cit⁺] phenotype is due to one or more mutations that result in, for example, the addition of a novel genetic regulatory element, gene-duplication with sequence divergence, or the gain of a new binding site, then it will be a noteworthy gain-of-FCT [Functional Coded elemenT] mutation.”

(Emphasis added.) Why did I bold that phrase? Because that’s exactly what happened.

First, there were a number of mutations in the genomes of bacteria in the A-3 line. At least some of these mutations were important in the later development of the Cit⁺ trait, as frozen bacteria that had the mutations were to varying degrees able to gain the ability when the experiment was rerun, but we don’t know which. This stage is referred to as the “potentiation phase,” and is not relevant to the bolded section.

After this, copies of the gene that allows the E. coli to metabolise citrate under anaerobic conditions were created in parts of the genome outside of the influence of the regulatory elements that prevented its use in aerobic conditions. This was the “actualisation” phase, and it’s important to realise that the strain created was “extremely weak.” To follow, then, was the “refinement” phase – further mutations were needed to let the new mutants all but take over their flask.

So, in other words, we have indeed got a case of “gene-duplication with sequence divergence.” If the Behe quote is correct then we have new functional coded elements!

It should go without saying, then, that in his own description of events Brian leaves out both the potentiation and refinement phases, which would pass the test laid out. Instead, he says:

So, the big question is: Did E. coli evolve into a Cit⁺ strain by natural selection? Or did mutations construct new and functional coded elements to its DNA? If so, it would be the first in recorded biological history. If not, then it would be just another loss or modification of a pre-existing piece.

This is a funny paragraph, as it doesn’t make a lot of sense. Is it an either/or condition, or a yes/no question? Natural selection is involved here, of course – you can’t ‘take over’ the environment without doing it. And as I said, by Behe’s definition we have new functional elements. So it’s both.

Brian’s flawed description of the process involved follows, phrased in a way that claims that the bacteria was “solving the problem,” which is not exactly an evidence-based claim on his part. He concludes his article:

So, the bacteria solved the problem of accessing an alternative food source by generating extra copies of the critical gene and by placing those copies under the control of an appropriate promoter. Does any of this resemble natural, undirected Darwinian evolution? Not at all. This amazing mechanism invented no new functional coded elements. It merely modified pre-existing elements.

Therefore, not only did the Cit⁺ bacteria not evolve in the molecules-to-man direction, but they showed what could only be ingenious DNA rearrangement mechanisms. What mainstream headlines portrayed as evidence for evolution is actually the opposite.

The article is actually a mix of arguments, as Brian does not seem to want to commit to fully either of them. It is apparently a result of design processes, but it’s also nothing major.

I wont even start on his strange definition of ‘evidence.’

---

*From the paper, if you want a better source than an off-hand comment from wikipedia:

Twelve populations of Escherichia coli have been propagated in the long-term evolution experiment (LTEE) for over 40,000 generations in a glucose-limited minimal medium. The medium also contains abundant citrate, which is present as a chelating agent, but E. coli cannot exploit citrate as a carbon and energy source in the well-aerated conditions of the experiment. The inability to grow aerobically on citrate is a long-recognized trait that, in part, defines E. colias a species. Spontaneous citrate-using (Cit⁺) mutants are extraordinarily rare, but a Cit⁺variant evolved in one population (designated Ara–3) around 31,000 generations.