# Fascinating findings from 21-year E. coli experiment.

## Recommended Posts

In 1988, an associate professor started growing cultures of
. Twenty-one years and 40,000 generations of bacteria later,
, who is now a professor of microbial ecology at Michigan State University, reveals new details about the differences between adaptive and random genetic changes during evolution.

##### Share on other sites

The key paper is this here, btw.

Proc Natl Acad Sci U S A. 2008 June 10; 105(23): 7899–7906.

Published online 2008 June 4. doi: 10.1073/pnas.0803151105.

Historical contingency and the evolution of a key innovation in an experimental population of Escherichia coli

Zachary D. Blount, Christina Z. Borland, and Richard E. Lenski

##### Share on other sites

Only 40,000 generations? I would have thought it would have been millions over that time span. Interesting results though.

##### Share on other sites

http://www.sciencedaily.com/releases/2009/10/091018141716.htm

"It's extra nice now to be able to show precisely how selection has changed the genomes of these bacteria, step by step over tens of thousands of generations," Lenski said.

Lenski's team periodically froze bacteria for later study, and technology has since developed to allow complete genetic sequencing. By the 20,000-generation midpoint, researchers discovered 45 mutations among surviving cells. Those mutations, according to Darwin's theory, should have conferred some advantage, and that's exactly what the researchers found.

The results "beautifully emphasize the succession of mutational events that allowed these organisms to climb toward higher and higher efficiency in their environment," noted Dominique Schneider, a molecular geneticist at the Université Joseph Fourier in Grenoble, France. <
>

E. coli cultures in the laboratory of Michigan State University evolutionary biologist Richard Lenski. (Credit: Greg Kohuth, Michigan State University)

• 1

##### Share on other sites

I find myself a bit confused with the terminology: "adaptive" and "random" genomic change.

Is there a significant difference that I'm missing, or are they classed simply as a change that offered an adaptive facet, and a change that did not?

##### Share on other sites

Could you point out the context in which they used it? I do not have the article with me right now.

But it is likely that you are on the right track.

##### Share on other sites
Could you point out the context in which they used it? I do not have the article with me right now.

But it is likely that you are on the right track.

The terms are not in the published paper, but in the Scientific American article that caught my attention above.

##### Share on other sites

This caught my eye.

Sequencing genomes of various generations of the bacteria, which had been frozen periodically over the years, Lenski and his team found that adaptive and random genomic changes don't necessarily follow the same patterns. Rather than a plodding equilibrium, even in a consistent environment, the interplay between these two kinds of genomic changes "is complex and can be counterintuitive," Lenski said in a prepared statement.
(bold supplied to assist in communication)

The failure to maintain "plodding equilibrium" is inconsistent with the commonly voiced assertion that stasis in "living fossils" like crocodilians can simply be attributed to an allegedly unchanging environment.

I wonder how quickly the implications of this discrepancy will spread from those who study bacteria to those who study croc history.

##### Share on other sites
I find myself a bit confused with the terminology: "adaptive" and "random" genomic change.

Is there a significant difference that I'm missing, or are they classed simply as a change that offered an adaptive facet, and a change that did not?

Hi Jill

There is evidence to suggest that when organisms are subject to environmental stress their mutation rate increases. If such a mutation is adaptive towards the stress then it is an 'Adaptive Mutation'.

Dougal

• 1

##### Share on other sites
Hi Jill

There is evidence to suggest that when organisms are subject to environmental stress their mutation rate increases. If such a mutation is adaptive towards the stress then it is an 'Adaptive Mutation'.

Dougal

Wow.

That's just amazing. And to think it may even happen with eukaryotes.

Thanks, dougalbod!

##### Share on other sites

Yes mutation rates vary, but adaptive mutation is not a point in the paper there was no selective pressure involved. In fact the increase in mutation rate is based on a frame shift mutation in the mutT gene. This gene is known to be involved in base transversions. Also it was only a minor element of the whole article and I am a bit surprised that SA highlighted it as much. The original article is this here, btw.

Genome evolution and adaptation in a long-term experiment with Escherichia coli.

Barrick JE, Yu DS, Yoon SH, Jeong H, Oh TK, Schneider D, Lenski RE, Kim JF.

Nature. 2009 Oct 18

In which Lenski and Kim are the corresponding authors.

Take a look at the abstract:

The relationship between rates of genomic evolution and organismal adaptation remains uncertain, despite considerable interest. The feasibility of obtaining genome sequences from experimentally evolving populations offers the opportunity to investigate this relationship with new precision. Here we sequence genomes sampled through 40,000 generations from a laboratory population of Escherichia coli. Although adaptation decelerated sharply, genomic evolution was nearly constant for 20,000 generations. Such clock-like regularity is usually viewed as the signature of neutral evolution, but several lines of evidence indicate that almost all of these mutations were beneficial. This same population later evolved an elevated mutation rate and accumulated hundreds of additional mutations dominated by a neutral signature. Thus, the coupling between genomic and adaptive evolution is complex and can be counterintuitive even in a constant environment. In particular, beneficial substitutions were surprisingly uniform over time, whereas neutral substitutions were highly variable.

The authors investigated two elements. The rate of genomic changes (=observable mutations) as well as the relative fitness increase of the new strains compared to the original parent strain.

What is important to keep in mind that even a mutation that gives a selective advantage may be lost due to stochastic events (drift). The larger the selective advantage the larger the chance that the mutation persists. Thus, in order for a mutation to spread through a population it first takes time for it to escape extinction. This time is inversely proportionate to the selective advantage it confers and the mutation rate with which this particular mutation occurs. And then the mutant spreads by selection. The required time for t to become the majority is again dependent on the selective advantage it confers.

--------Just running out of time, if interest exists I will continue when I get a break sometime-----------------------

Just btw. increasing mutation rate is for most multicellular organisms a bad idea. Think cancer. Asexually reproducing organisms can do it, as even if a lot of sister cells die, almost complete copies of their gene sets will survive (in those that adapt). This strategy obviously does not work with other organisms as well...

##### Share on other sites

Thanks CharonY. If you care to continue I assure you it will be appreciated. ==

##### Share on other sites

Ok continuing the post:

So keep in mind that the authors took a look at fitness increase and mutation rate (=genomic changes). One question is what the expected relationship between those two is.

To summarize what I said in the earlier post the time until a beneficial mutation escapes extinction is $\omega$ $\approx$ 1/(2SN$\nu$)

And the time for the beneficial mutation to spread to 50% of the population (and assuming a Poisson distribution of offspring) is

$\tau$ $\approx$ log2(0.5N)/S

With S being the selection coefficient (high values indicate higher selective advantage), N being population size and $\nu$ the rate of beneficial mutations.

The important bit is only to keep the relation of the parameters in mind.

So assume the following scenarios:

1) no effect on the selective advantage or beneficial mutation rate due to substitutions of beneficial mutations. In this case both, the rate of genomic changes as well a increase in fitness should remain constant.

2) assume that the number of beneficial mutation sites is finite. The more beneficial mutations it already has, the less likely it is, to gain a new one. Hence $\nu$ declines over time. Thus, the wait time $\tau$ becomes longer and both, the fitness increase, as well as the genomic change rate will decelerate over time.

3) assume that the selective advantages decline as more beneficial mutations arise. I.e. the first mutations gives a large boost in selective advantages, but the second one only adds a little bit more to it, etc. With a decline of S both $\omega$ and $\tau$ increase. And the effect would be a deceleration of fitness increase and genomic change rate.

Note that in all scenarios fitness increase as well as genomic change rates either decline or remain constant, without the need of changing the mechanisms that lead to mutation. In scenario 2, for instance the deceleration is merely due to the constraints on targets that may mutate (i.e. due to the limited beneficial sites).

Now take a look at the abstract again.

Although adaptation decelerated sharply, genomic evolution was nearly constant for 20,000 generations.
. Here they state that the fitness increase decelerated, whereas the genomic change rate was constant. This does not fit any of the simple models stated above. How do the authors explain this?

Stay tuned for more after a short break.

##### Share on other sites
Stay tuned for more after a short break.
Eeee!

This is fun! Thanks for helping me grock this.

##### Share on other sites

Just a short note: I messed up on a point

2) assume that the number of beneficial mutation sites is finite. The more beneficial mutations it already has, the less likely it is, to gain a new one. Hence $\nu$ declines over time. Thus, the wait time $\tau$ becomes longer and both, the fitness increase, as well as the genomic change rate will decelerate over time.

The wait time is of course $\omega$. That is what you get when posting in a rush. Geez.

##### Share on other sites

So much for short. Anyway just to continue the thread:

The easiest explanation for the discrepancy is that the majority of mutations are not beneficial. The earlier increase of fitness is possibly due to rapid adaptation to the new medium, but after that the continuing mutations will be (near) neutral. The initial beneficial burst would then be overshadowed by the continuing accumulation of neutral mutations.

However the authors found several aspects that would contradict this scenario. First, if the majority were neutral mutations, one would expect a higher frequency of synonymous mutations (i.e. mutations that do not translate into a different amino acid). In fact, the authors found that all mutations in coding regions were in fact non-synonymous.

Second, if the spread was due to random drift then one would expect different mutations in the different cultivations. However, in 11 independent cell lines a high proportion of similarities were found.

Third, if drift was a major factor (as opposed to selection) then many occuring mutations are unlikely to be fixed in the population. However, mutations were found to be persisting through time (they froze sample throughout the generations to be able to track changes).

And finally they generated strains artificially with the same mutations and found that at least8 loci actually did confer fintess advantages (and were therefore not neutral).

##### Share on other sites

Such an amazing experiment. I personally don't think I could have the patience to conduct an experiment over that length of time so to me it's all the more amazing.

Plus I like it because it shoots another hole in the IDist's very non scientific claims.