help with Richard Lenski's e-coli experiment

[talkie toaster]

I'm not a biologist or microbiologist so any replies I'd be grateful if you could make it the big letter, A is for Apple version.

The experiment is a beautiful example of science in action and the process of evolution.

However, someone has brought to my attention Plasmids and used it to indicate that e-coli's ability to process citrate in the experiment counter acts any idea of mutating genes and suchlike, namely it takes an unused gene from it's plasmid and uses that to alter it's ability of citrate processing which it then passes onto other e-coli.

So, they said, it wasn't mutation followed by natural selection because the gene was already there in it's associated plasmid.

This sounds like crap and any microbiologist worth his lab coat would know about this.

I've never heard of Plasmids and searching the internetwebbythingy has provided stuff I don't understand.[url=http://www.freesmileys.org/smileys.php][/URL]

Does anybody know anything about plasmids and can they do this.

Thanks .

 

[geni]

The problem with that line of argument would be:

1)no evidence it was the case
2)Most the strains didn't develop the ability to use citrate
3)the early citrate using mutants were not very good at using it.

 

[blutoski]

This is a boilerplate (and arguably valid) rebuttle to previous similar experiments that may have claimed they were 'proof' of evolution: a change in allele frequencies is not entirely convincing without demonstration of a de novo beneficial allele mutation.

That's what makes Lenski's experiment meaningful: this allele was not in a plasmid. It is a new allele formed from mutations in a gene sequence over time, which turned out to be beneficial.

Background: plasmids are real of course, and they are pretty much as described by the IDer... typically, the allele on a plasmid was somewhere in the population already, and just gets distributed (from bacterium to bacterium) and amplified (bacteria with the plasmid reproduce more succesfully) when the culture's environment changes to require it. You can understand why this would be weak sauce as 'proof' of the evolutionary hypothesis. It's not a demonstration that new properties can appear.

Lenski's experiment specifically shows a new allele - a new property. This sequence was not in any plasmid previously. Further: we can observe the stepwise mutations that led to the new gene sequence.

It's a demonstration of an extremely well documented beneficial de novo mutation in vitro.

It is not the only de novo beneficial mutation recorded... just the first that was produced intentionally, and with such complete documentation.

 

[Rolfe]

Point of in formation. It's Escherichia coli, abbreviated E. coli.

Capital E, full stop, space, lower-case coli. Correctly, it should be italicised.

/pedant.

Rolfe.

 

[ben m]

I'm not a biologist or microbiologist so any replies I'd be grateful if you could make it the big letter, A is for Apple version.

The experiment is a beautiful example of science in action and the process of evolution.

However, someone has brought to my attention Plasmids and used it to indicate that e-coli's ability to process citrate in the experiment counter acts any idea of mutating genes and suchlike, namely it takes an unused gene from it's plasmid and uses that to alter it's ability of citrate processing which it then passes onto other e-coli.

First: Lenski et. al. are quite literally watching the genome. They've got monthly-ish samples of every population in a deep freezer, and they pull out these frozen ancestors and sequence them. You can see the mutations happening; they have records showing where Flask 1991-10-G had an ATGG and Flask 1991-11-G had an ATGA instead. You can see the "neutral" mutations as well as the active ones. I THINK (but I don't know the lit that well) they know what mutation led to the citrate thing (Has anyone read http://symposium.cshlp.org/content/early/2009/09/22/sqb.2009.74.018.full.pdf+html ? ) and that it's a smallish change in the regular genome, not the sudden appearance of a new gene out of nowhere.

Second: one of Lenski's most dramatic results was when he showed (before sequencing anything) that his various cultures were *different*, and that they picked up new differences at different times.

None of his cultures at the 30,000 generation stage ("30K") had any cit+ ability. He replayed the experiment from the 30,000 generation stage and none of them evolved the rare ability. Ditto for the 31K generation---no cit+, no tendency to develop cit+. Ditto for the 31,500 generation---no cit+, no tendency to evolve cit+.

The Ara-3 32K generation, despite being directly descended from the 31K generation, was different. It had no cit+, but when this sample was restarted it would evolve cit+ with reasonably high probability.

If you want to explain this with a plasmid---remember that plasmids are inherited. If the 32K generation had some ready-to-go cit+ plasmid waiting to be spliced in ... well, it must have inherited that plasmid from its parents, which means the 31K generation must have had the same plasmid. But it didn't---the 32K generation had something that the 31K simply didn't have. Mutations can do that. Plasmids (or any other form of "secret intelligently-designed apparatus being carried around somewhere") cannot.

 

[noreligion]

I'm not a biologist or microbiologist so any replies I'd be grateful if you could make it the big letter, A is for Apple version.

The experiment is a beautiful example of science in action and the process of evolution.

However, someone has brought to my attention Plasmids and used it to indicate that e-coli's ability to process citrate in the experiment counter acts any idea of mutating genes and suchlike, namely it takes an unused gene from it's plasmid and uses that to alter it's ability of citrate processing which it then passes onto other e-coli.

So, they said, it wasn't mutation followed by natural selection because the gene was already there in it's associated plasmid.

This sounds like crap and any microbiologist worth his lab coat would know about this.

I've never heard of Plasmids and searching the internetwebbythingy has provided stuff I don't understand.[url=http://www.freesmileys.org/smileys.php][qimg]http://www.freesmileys.org/smileys/smiley-basic/dunce.gif[/qimg][/URL]

Does anybody know anything about plasmids and can they do this.

Thanks .

That someone is being entirely dishonest and quite possibly intentionally deceptive. The ability to metabolize citrate is one that used two mutations. The "someone" is most probably an IDiot since this one aspect of the ongoing experiment has shown that the ID concept of irreducible complexity is nothing more than made up mumbo jumbo to impress the uneducated.

 

[Soapy Sam]

Bacteria do exchange DNA via plasmids. This is commonplace. Is someone questioning it?

 

[NewtonTrino]

I just wanted to say I really appreciate all of the people who responded in such detail. I'm fairly familiar with the experiment and it's really great work but you guys were able to totally smash the premise in the OP quite quickly and convincingly.

 

[Lukraak_Sisser]

In awnser to the OP.
Plasmids is a name used in molecular biology for small circular segements of DNA capable of autonomous replication in a host organism. In essence they are mini genomes. Some of them originate from viruses, some of them are just part of the way organisms work. And given how useful they can be, there is a vast number of artificial plasmids out there now.
However, since replicating a plasmid takes energy, its a slight disadvantage in replication for E.coli.
Plasmids are not stable, they can be lost on replication, so over time only plasmids that give some growth advantage are maintained.

I suspect that the author the OP mentioned assumes a plasmid with cit+ has entered the population. There are a number of errors with that hypothesis though
First, no plasmids with cit+ are known to exist
Second, introducing plasmids into E.coli generally involves active effort of a scientist so even IF an unknown plasmid had somehow contaminated one of the flasks it would not enter the population
Third, once a plasmid is in its genes are active in E.coli, and the gene would not quickly hop into the genome (it can happen, just not fast) so a large number of the cultures would have plasmid contamination, which is immediatly appearant once they isolate DNA
Fourth, as has already been pointed out, the Lensky's group has tracked the relevant mutations by sequencing and there is no indication of gene insertion, just point mutations.

So in short, it seems like a lazy and badly researched denial by someone who just cannot accept nature the way it is. They would have been better off going with the micro-evolution thing or ignoring it.

 

[talkie toaster]

Many thanks for the replies.

BTW, in my OP I wasn't using the "someone suggested to me......" ruse of posting in order to cover up the fact that "someone" is actually me.

Not knowing what Plasmids were or how they work or indeed the full process of the Lenski experiment it seemed a possible outcome for my given knowledge.

However, as the saying goes "you're never too old to learn.." so thanks for increasing my understanding.

I guess using examples of things that prove a point without fully being cognisant of all the details can cause problems when questioned with something that is also short on details regarding the opposing view.

 

[bokonon]

Further: we can observe the stepwise mutations that led to the new gene sequence.

In theory, we can, since Lenski has archived samples every 500 generations.

In practice, I don't know that we have. Lenski's 2008 paper states:

To find the relevant mutations, we will perform whole-genome resequencing, which has become a powerful approach that is well suited to experimental evolution (57–59). We expect to find dozens of mutations relative to the ancestor (22), which will complicate identification of those changes that were important specifically for the origin of the Cit function.

If such sequencing has been completed, I have not seen the results, and would appreciate a link if they are available.

First: Lenski et. al. are quite literally watching the genome. They've got monthly-ish samples of every population in a deep freezer, and they pull out these frozen ancestors and sequence them. You can see the mutations happening; they have records showing where Flask 1991-10-G had an ATGG and Flask 1991-11-G had an ATGA instead.

If these records have been published, I'd like to know where.

The ability to metabolize citrate is one that used two mutations.

Not really. E. coli has the ability to metabolize citrate, what it lacks is the ability to move the citrate from its outside to its inside if oxygen is present. Lenski's paper speculates that three mutations were involved in the development of this ability in these samples:

First, a "potentiating" mutation, which predisposed the cells to be able to generate the citrate-transporting protein.

Second, a weak ability to transport citrate.

Third, a strengthening of this ability.

The "someone" is most probably an IDiot since this one aspect of the ongoing experiment has shown that the ID concept of irreducible complexity is nothing more than made up mumbo jumbo to impress the uneducated.

Nonsense. This experiment does not address the notion of irreducible complexity. A single protein can create a channel for the movement of citrate from the outside to the inside of the bacteria, so in theory a single mutation could generate it.

In practice, it's even simpler: E. coli can already create such a channel if oxygen is not present, so a single mutation which permitted this to happen in the presence of oxygen would theoretically suffice.

Bacteria do exchange DNA via plasmids. This is commonplace. Is someone questioning it?

I believe the bacteria chosen for this long-term experiment are asexual, which I interpret to mean that they're not swapping plasmids. I could be wrong.

First, no plasmids with cit+ are known to exist

Not true. Lenski's paper mentions them:

Indeed, atypical E. coli that grow aerobically on citrate (Cit) have been isolated from agricultural and clinical settings, and were found to harbor plasmids, presumably acquired from other species, that encode citrate transporters (44, 45).

Fourth, as has already been pointed out, the Lensky's group has tracked the relevant mutations by sequencing and there is no indication of gene insertion, just point mutations.

Everybody seems to be assuming this is the case, but WHERE IS THE DATA? It's been two years since the initial paper, so it's certainly possible that the sequencing has been done, and the relevant genetic changes have been identified, but I've looked for it and I haven't found it.

 

[Matt the Poet]

Plasmids are not stable, they can be lost on replication, so over time only plasmids that give some growth advantage are maintained.

Not quite true. Most plasmids have evolved stabilisation systems that ensure they’re maintained even in the absence of a growth advantage – either by making sure they get physically segregated during division or through expressing a stable toxin and a less stable antitoxin that effectively ‘addicts’ the bacterium to the plasmid.

Interestingly, the latter appears to turn up in bacterial genomes as well, where it is sensitive to poor nutritional conditions – in other words if there’s not much food around you get a sort of ‘bacterial apoptosis’ or cell suicide as toxin levels build up.

Sorry, but it’s so rarely that a thread comes along that actually nails a personal specialty (before I quit science I did a PhD in this stuff) that I felt the urge to share.

Anyhoo, it’s the work of about an hour to check for plasmid contamination in a bacterial population. Take out all the DNA, run it through agar gel for half an hour at the relevant voltage and, if I recall correctly (I was doing this a good ten years ago) the moderately sized plasmids will migrate far enough to be seen while the DNA hangs about in the well at the top. Lenski has a background in plasmid evolution, it seems unlikely that they wouldn’t have looked for this possibility.

 

[ben m]

Not quite true. Most plasmids have evolved stabilisation systems that ensure they’re maintained even in the absence of a growth advantage – either by making sure they get physically segregated during division or through expressing a stable toxin and a less stable antitoxin that effectively ‘addicts’ the bacterium to the plasmid.

I will never cease to be amazed.

 

[ben m]

Anyhoo, it’s the work of about an hour to check for plasmid contamination in a bacterial population. Take out all the DNA, run it through agar gel for half an hour at the relevant voltage and, if I recall correctly (I was doing this a good ten years ago) the moderately sized plasmids will migrate far enough to be seen while the DNA hangs about in the well at the top. Lenski has a background in plasmid evolution, it seems unlikely that they wouldn’t have looked for this possibility.

I suppose you could argue that the parent population had this cit+ plasmid, but in a way that you wouldn't see on a gel. Suppose the plasmid (a) had an extremely low penetrance, with a bare handful of copies of it in the whole beaker, and (b) was not actually expressing anything in plasmid form.

Then in generation 31500 (or whatever) one of the rare individuals with the plasmid accidentally spliced it into the genome and separated from its inhibitor (or whatever). That individual became cit+ and started taking over the colony. Voila! You start with a beaker with no visible plasmids and no cit+, and you end up with a beaker with no visible plasmids and cit+. And if it happens this way you can still imagine that the information part of the deal---the actual code for a functioning citrate transport pore---was Designed By God and waiting around in the population.

Just sayin'. I'm not sure you'd be able to rule this out without (a) common sense---neutral drift would erode any information God stored this way---and (b) sequencing, which they did.

 

[bokonon]

I believe the bacteria chosen for this long-term experiment are asexual, which I interpret to mean that they're not swapping plasmids.

It's actually stronger than that. According to an earlier paper, these bacteria began the experiment with no plasmids to swap:

The derivation of the genotypes used in this study has been previously described (LENSKI et al. 1991). Briefly, all strains were derived from a single genotype of E. coli B. The ancestral genotype does not contain any plasmids or functional phage, and therefore its evolution is strictly clonal.

I'm still waiting for one of the folks claiming that the sequences are available to produce evidence supporting that claim...

 

[bokonon]

One paper which deals with the plasmid-derived transport protein in E. coli identifies it as being composed of 487 amino acids:

The open reading frame located at 13.9 min on the E. coli chromosome (designated citT), starting 50 bp downstream of the citG stop codon, encoded a protein of 487 amino acids

(citT is the identifier given for citrate Transport). A protein composed of 487 amino acids would require over 1400 base pairs of DNA to code.

Consider for a moment how a 1400-base-pair gene might arise.

It might happen that 1399 of those base pairs were just coincidentally doing something else, and the single change that made it possible for this protein to do something new (move citrate across the membrane) didn't break what it was doing before, or DID break it, but what it was doing before wasn't essential, at least once it could move citrate. Or 1398 and 2 changes, or 1370 and 30 changes, or 1260 and 140 changes, etc.

It might happen that all the DNA to code the protein was in place before, and the bacteria just found a novel way to splice those pieces together before it was translated into a string of amino acids.

A big chunk of new DNA might have come from somewhere else -- a plasmid, or a bacteriophage. I'm confident that plasmid contamination would have been detected. Viral contamination, I'm not so sure about, which is why I'd like to know that the sequences were available and had been scrutinized by experts in the subject.

This 1982 paper describes an earlier instance of "Chromosomal Mutation for Citrate Utilization," and hypothesizes that two distinct mutations (one to citA, and one to citB) could do it, without really confirming that this is what happened. However, the author did conclude that

Several observations suggested that citrate utilization did not arise from a simple point mutation.

At this point, we (I) don't really know whether the change observed in Lenski's experiment was similar to a "plasmid-style" citT mutation, or a "citA/citB" mutation, or some completely new solution to the citrate transport problem. Where's the sequence data? Beuhler? Anybody?

 

[bokonon]

I wrote to Richard Lenski, and he confirms that the sequencing is still a work in progress,

with many entire genomes being sequenced and analyzed to understand the genetic bases of that evolutionary transition. And as we gather such information, we must perform additional (and often difficult) experiments to confirm the relevance of the various mutations that we find.

All the bacteria in a dish of E. coli don't have a single uniform genome, even if they all came from a single cell. It's to be expected that the process of winnowing significant changes from background noise will be tedious and time-consuming. I look forward to the time when we DO know whether this population engineered a novel solution to the citrate transport problem, or re-discovered a known method -- and, even more interesting, the steps along the way.