Uncommon Descent Serving The Intelligent Design Community

Evolution of an Irreducibly Complex System – Lenski’s E. Coli

Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

On another thread we have been discussing abiogenesis in particular, but there was also some discussion about the evolution of an irreducibly complex system. Commenter CHartsil indicated that “we actually watched an IC system evolve” in reference to Lenski’s E. coli research. At my request, he has posted a brief summary of the research and his take, which I am now elevating to a new thread on this important topic.

For those who disagree with CHartsil’s take, strong objections on substantive grounds are of course welcome, whether relating to Lenski’s research or CHartsil’s interpretation of it, but not irrelevant personal attacks. Thank you.

—–

Guest Post by CHartsil:

This is a pro-ID board so I doubt I need to explain irreducible complexity. When arguing against it, most will bring up Ken Miller or Nick Matzke. They have great points but theirs are indirect and theoretical pathways for systems considered IC. That’s why I’m fond of Lenski’s cit* E. coli.

This particular strain of E. coli evolved the ability to metabolize citrate aerobically. While most E. coli can do this anaerobically, part of the definition of wild-type E. coli is actually the inablity to use citrate as a substrate aerobically. This may not have been a terribly fascinating addition of function if not for the frozen fossil records kept by Lenski et al.

These frozen generations allowed Lenski to determine that this trait was not acquired via a single mutation as it could only be repeated after generation 20,000. Given the distinct cladistic division amongst the populations at the border generation, it was determined that there were at least two potentiating mutations prior to the cit* event.

In this third clade a tandem duplication resulting in a novel regulatory module leading to the aerobic cit* could be repeated and verified. It has been noted since that the fitness of the population has been improving without notable upper limit, increasing based on the number of copies of the new regulatory module.

I find this to be sufficient in warranting the dismissal of the concept irreducible complexity. In Lenski’s E. coli, we observe the rise of a new function resulting from a new gene and new gene regulation. This function is comprised of now interdependent components which demonstrably did not exist in parent generations. It is by definition irreducibly complex and it was observed to evolve.

—–

Nota bene: for purposes of the above discussion, CHartsil is using the following definition of irreducible complexity: “a system comprised of interdependent parts, the removal of any of which will cause the system to cease functioning.”

Comments
Follow up #151,
Martincorena et al: Our observations suggest that purifying selection has driven the evolution of the local point mutation rate in E. coli to reduce the risk of deleterious mutations. This contrasts with most earlier theoretical work that proposed variants of bet-hedging in which frequent positive selection in changing environments leads to the emergence of hypermutators1–3. Instead our observations are in line with an evolutionary risk-management strategy26 in which sustained stronger purifying selection at specific genes favours individuals with preferential protection or repair at these loci, even at a cost of reduced protection of other genes (see Supplementary Information, section 4.3, for a detailed description of the model). In this way, the rate of deleterious mutations in the genome can be efficiently reduced without excessive investment in protection or repair. In addition, this could increase the rate of nondeleterious mutations, so raising the adaptive potential of the population in case of an environmental change. We can only speculate about the molecular mechanisms underlying the localized reduction in spontaneous mutations.
Box
February 27, 2015
February
02
Feb
27
27
2015
01:23 PM
1
01
23
PM
PDT
wd400:
The medium has a small amoung of glucose. They can survive absent a way to use citrate, than can prosper with a way to use it.
From the Wikipedia article on Lenski's experiment, the bacteria were grown on a minimal growth medium. I suspect that, though "minimal," there was still plenty of glucose available. Hence, the cit+ mutations were simply, shall we say, a 'bonus.' So I still don't see how this could be considered IC.
I don’t know how you think drift would find it (or even how that makes any sense at all, since a mutation that adds to efficient use of citrate will have a selective advantage)
From the Wikipedia page on Lenski's experiment:
The authors interpret these results as indicating that the evolution of citrate use in this one population depended on one or more earlier, possibly nonadaptive "potentiating" mutations that had the effect of increasing the rate of mutation to an accessible level.
There was likely NO selective advantage here because clones taken from before the 20,000th generation could not develop cit+, and the cit+ didn't develop until the 31,000th generation; hence, as the authors say, they were "possibly nonadaptive." (In fact, they were likely "nonadaptive." The fitness did not go up substantially until the cit+ mutation appeared [duplication]) As to a new definition of evolution, why use the word "evolution" when 'change' is just as good? Evolution implies some progressive movement. What "progress" has been made? Is E. Coli now not E. Coli? Should we call it something else? The growth medium was "minimal". What if they grew the Cit+ populations on a glucose rich medium, then what would happen? I suspect the E. Coli would lose the cit+ ability and go back to wild-type E. Coli. So, then, we're back where we started from. And this is evolution? Why not call it what they call it at Wikipedia: "adaptation"? Here's the Webster's Dictionary definition of "adaptation": a change in a plant or animal that makes it better able to live in a particular place or situation. So, I'll stick with the word "change." From Wiki:
The researchers also found that all Cit+ clones sequenced had in their genomes a duplication mutation of 2933 base pairs that involved the gene for the citrate transporter protein used in anaerobic growth on citrate, citT. The duplication is tandem, resulting in two copies that are head-to-tail with respect to each other. This duplication immediately conferred the Cit+ trait by creating a new regulatory module in which the normally silent citT gene is placed under the control of a promoter for an adjacent gene called rnk. The new promoter activates expression of the citrate transporter when oxygen is present, and thereby enabling aerobic growth on citrate.
There wasn't a 'new' gene that developed, there was an already present gene that was duplicated and put under the regulation of another module. The "potentiating" mutations quoted above had the effect of "increasing" the mutation rate. Everything that happened looks to me to have happened in some kind of a relatively 'ordered' way. It reminds of Shapiro's NGE. I suspect that when we know more, we'll find out that bacteria, when cultured for so long in a somewhat restricted feeding environment will turn on certain mechanisms that allow certain areas of the genome to start moving relevant parts of the genome around. Then, in a random way, something that increases the fitness will be found. But this smacks of "engineering," as in NGE.PaV
February 27, 2015
February
02
Feb
27
27
2015
01:12 PM
1
01
12
PM
PDT
The conjecture is that mutations rates can be tweaked in different genomic regions, not that mutations in a given region could be non-random with respect to fitness.wd400
February 27, 2015
February
02
Feb
27
27
2015
01:10 PM
1
01
10
PM
PDT
WD400: Confirming this was the paper you were taking about, since it does nothing to show mutations are not random with respect to fitness.
I have to say that this is not quite true. Martincorena et al argue for non-random mutations with regard to fitness. Although it is also true that they sugarcoat their message with political correctness and extreme caution.
Thus, although mutations are generally assumed to occur independently of their fitness effect, it is conceivable that local mutation rates themselves might evolve, resulting in genomes whose mutations occur non-randomly: more frequently where they are more often advantageous and less frequently where they are most deleterious. Currently, however, there is little evidence that local mutation rates have been optimized during evolution, with the limited exceptions of bacterial contingency loci and somatic hypermutation in the vertebrate immune system. Our understanding of how the mutation rate varies along a genome has been restricted by the lack of reliable approaches to measure local mutation rates on a large scale. Experimentally, absolute mutation rates can be determined using gene reporters in fluctuation tests, but these are unsuitable for measurements in native genes. Alternatively, in theory,relative mutation rates can be estimated from the accumulation of mutations at selectively neutral positions. Indeed, synonymous or non-coding sites are often used as proxies 7,8 and intriguing correlations between local synonymous substitution rates and gene function or fitness cost have been reported 8,9. However, because selection can act on these sites through factors like codon-usage preference, RNA-folding stability and cis-regulatory elements, interpretation of these observations in support of optimized mutation rates has remained contentious.
Box
February 27, 2015
February
02
Feb
27
27
2015
01:03 PM
1
01
03
PM
PDT
I guess you missed my point. If the organism “must” use citrate in an anaerobic environment, and it does not have this putative IC system, then how could it possibly survive??
The medium has a small amoung of glucose. They can survive absent a way to use citrate, than can prosper with a way to use it. That put's pay to most of the rest, but this passage seems to invent a new meaning for the word evolve.
his seems to mean that the ‘added’ mutations simply enhance this reduced capacity. Given enough time, we would expect that through neutral drift, and perhaps some kinds of directed mechanisms, a much greater ability to process the citrate would develop (not ‘evolve’)
A change in genotype leading a change in phenotype is evolution. I don't know how you think drift would find it (or even how that makes any sense at all, since a mutation that adds to efficient use of citrate will have a selective advantage)wd400
February 27, 2015
February
02
Feb
27
27
2015
12:24 PM
12
12
24
PM
PDT
Link was to here: http://www.ncbi.nlm.nih.gov/pubmed/22522932 Confirming this was the paper you were taking about, since it does nothing to show mutations are not rando with respect to fitness.wd400
February 27, 2015
February
02
Feb
27
27
2015
12:12 PM
12
12
12
PM
PDT
WD400 I did miss it. Apologies your link is not working please repost.Andre
February 27, 2015
February
02
Feb
27
27
2015
12:04 PM
12
12
04
PM
PDT
Hey Andre, did you miss my reply in #70? (I assume, since you said this "I cited a paper that shows that mutations are not random with good evidence. What do you do? You ignore it…. why? Does it challenge your religious beliefs?" that you weren't simply ignoring it)wd400
February 27, 2015
February
02
Feb
27
27
2015
11:57 AM
11
11
57
AM
PDT
I apologise this is off topic but news maybe you want to look into this? http://www.nature.com/nature/journal/vaop/ncurrent/full/nature14172.htmlAndre
February 27, 2015
February
02
Feb
27
27
2015
10:48 AM
10
10
48
AM
PDT
CHartsil:
PaV: “What exactly was “removed” from the system?” Response: The potentiating mutations prior to the 20,000th generation. Also, if either the duplicate gene or the new regulator is removed, the cit* ceases functioning.
How do you know they were removed?
PaV: “if the bacteria “cease[d] functioning,” then how did it “evolve”? That is, if it ceased functioning, then it would die. Dead things don’t “evolve.”” Response: It’s talking about individual systems, not the organisms in which the systems reside.
I guess you missed my point. If the organism "must" use citrate in an anaerobic environment, and it does not have this putative IC system, then how could it possibly survive?? If you think it through, the answer is likely that the bacteria has a very reduced ability to synthesize the citrate; but, nonetheless, a reduced ability is there, otherwise it could not survive. The logical conclusion is that the ones that "survive," survive because they already have some kind of reduced ability to synthesize the citrate. This seems to mean that the 'added' mutations simply enhance this reduced capacity. Given enough time, we would expect that through neutral drift, and perhaps some kinds of directed mechanisms, a much greater ability to process the citrate would develop (not 'evolve').PaV
February 27, 2015
February
02
Feb
27
27
2015
10:22 AM
10
10
22
AM
PDT
Joe,
If you need 5 parts and all you have are two, you lose.
Why? You start with a primitive function, add more parts, and get a more complex system with new secondary functions (exaptations) which may be selected for and fine-tuned.Piotr
February 27, 2015
February
02
Feb
27
27
2015
09:17 AM
9
09
17
AM
PDT
H, I am not particularly arguing about particular cases, I have noted objections and have identified that IC is both real and relevant. That is a baseline, it really has to be faced. I am not satisfied that ever so many objectors are willing to admit even that much. Same, for the deeper challenge, FSCO/I. Multiple, matched interacting parts that all have to be present then properly arranged and coupled to achieve function. Function, that if just one of the multiple core parts is missing, will fail. The claimed counter example is a case of two components to achieve a function, not several. E Coli already can address citric acid, it already has promoters, just the TWO need to come together. For me, the first interesting issue would be how do we get to the two functions and to the mutual fit, the real innovation. The coupling of the two components is then in effect one join, and it seems hard to find . . . significant. That, to my mind, does not seem to be a test of what IC is about. But, as here we have A + B --> AB, that will be trumpeted as dismissing the real issue. That rather reeks of a strawman to me, with all due respect. KFkairosfocus
February 27, 2015
February
02
Feb
27
27
2015
08:36 AM
8
08
36
AM
PDT
Me: Nearly all the proteins in the flagellar assembly have homology with other cellular proteins Joe: So what?
So it substantially increase the likelihood that serial additions can occur. If every component has to be invented afresh, that's a much bigger problem.
You need the correct number of subunits. You need the correct configuration of those parts and you need some way to communicate with it so it can be controlled.
For the Full Monty modern E Coli version, perhaps. But this is a derived state, where the modern parts have been subject to a long history of the co-evolutionary process outlined above. Each amendment causes tuning of the existing components in view of the overall interaction. If A alone had a function, then B is added, A evolves in the presence of B, to the point where B can no longer be removed.Hangonasec
February 27, 2015
February
02
Feb
27
27
2015
08:33 AM
8
08
33
AM
PDT
Hangonasec: Who said anything about ‘need’? Joe: Organisms have needs. They need functioning systems and subsystems.
And if they are already viable and competent, those immediate needs are met, by iterated replication of the current blueprint. You are talking about the 'need' for an alteration, and then inventing some kind of Universal Block to that need ever being met. One change is OK? Right, evolutionist, try 2! 3! 40! How do ya like them apples!
Me: If you have a 2 part structure, another part can be added to it. Joe: maybe and maybe not. It all depends. If you need 5 parts and all you have are two, you lose.
So, if that is what you really 'need', you go extinct. Who cares? The space of possible genomes is littered with places evolution cannot go, or cannot stay long when the environment changes. This does not define the entire space. Modern life is populated exclusively by lineages that did not go extinct. If there was an unfulfillable 'need' in some other lineage, it has been erased from the page and need not trouble us. All one has to do is assert that there was not such an unfulfillable need in ours. That is, at least, a perfectly valid logical possibility. You can make the counter-assertion that there was, but the evidence is against you - here we are!Hangonasec
February 27, 2015
February
02
Feb
27
27
2015
08:20 AM
8
08
20
AM
PDT
Having a random distribution does not mean they are all accidents, errors and mistakes.
We showed where irreducible complexity can occur in biology through two different processes.
You don't even know what you are trying to demonstrate.Joe
February 27, 2015
February
02
Feb
27
27
2015
06:23 AM
6
06
23
AM
PDT
nightlight: Intelligent actions in my semantics are those which on average improve survival odds (or for general kind of system, improve relevant ‘rewards – punishments’) compared to the actions which are picked randomly out of all possible actions. Has to be better than a random distribution, then. There's no evidence that genetic mutations are intelligent. fifthmonarchyman: The argument has always been that an IC system can not be the result of evolution not that they can’t arise by other natural processes. "Can not" is a very strong claim. We showed where irreducible complexity can occur in biology through two different processes.Zachriel
February 27, 2015
February
02
Feb
27
27
2015
06:18 AM
6
06
18
AM
PDT
Rivers do not evolve as evolution requires reproduction.Joe
February 27, 2015
February
02
Feb
27
27
2015
06:09 AM
6
06
09
AM
PDT
Hangonasec:
Nearly all the proteins in the flagellar assembly have homology with other cellular proteins
So what? You need the correct number of subunits. You need the correct configuration of those parts and you need some way to communicate with it so it can be controlled.Joe
February 27, 2015
February
02
Feb
27
27
2015
06:08 AM
6
06
08
AM
PDT
fifthmonarchyman: How do you think lakes arise? All sorts of ways. As rivers and landscapes evolve, lakes can occur due to landslides, oxbows, sinkholes, tectonic forces, glaciers, even volcanoes. fifthmonarchyman: Lakes are the result of the combination of two separate components a river and an obstruction/dam. There is no river without geology. fifthmonarchyman: It’s really difficult to have a discussion with someone who attributes every conceivable change to evolution. As we stated, the evolution of rivers is not the same thing as biological evolution. You're the one who introduced rivers, after all, claiming "we would never say that a reservoir evolved from a precursor river." But this has nothing to do with biological evolution. Not sure why you brought it up. fifthmonarchyman: Scaffolding is not evolution Scaffolding in biology can occur when a structure undergoes duplication and mutation, then parts are knocked out. So you might have A then A-A then A-A-A1 then A1-A2-A3 then A1-A3; wherein the complex A1-A3 can't evolve directly, but is irreducible. A2 is the scaffold. fifthmonarchyman: Addition and specialization are not evolution. Of course it is. For example, we have A, then a helper is added B. Then some of the function migrates from A to B, so we have A1-B1; wherein the complex A1-B1 is irreducible. fifthmonarchyman: Can you name a single natural process that you would not consider to be “evolution”? If you mean biological evolution, then anything not biological.Zachriel
February 27, 2015
February
02
Feb
27
27
2015
06:07 AM
6
06
07
AM
PDT
Hangonasec:
Who said anything about ‘need’?
Organisms have needs. They need functioning systems and subsystems.
If you have a 2 part structure, another part can be added to it.
maybe and maybe not. It all depends. If you need 5 parts and all you have are two, you lose.Joe
February 27, 2015
February
02
Feb
27
27
2015
06:06 AM
6
06
06
AM
PDT
KF - if one is arguing for a particular case declared IC, it is a diversion to demand how some component or other came into existence in the first place. The existence of competent ATPase activity can be taken for granted in all cellular life. How it got there is not a question of direct relevance to the question of whether flagella are or are not IC, in the sense of being 'unevolvable' from a non-IC precursor state.Hangonasec
February 27, 2015
February
02
Feb
27
27
2015
05:50 AM
5
05
50
AM
PDT
H, I have to go now, on of those doctor wait type days -- long wait, painful, expensive, and you are out of your own control. With bureau -rats on the other sides of the desks, likely hostile. The issue is simultaneity of multiple well matched parts as a pivotal issue. As Menuge highlighted. And BTW, without a viable mechanism for ATP energy battery molecules or the equivalent no cell, then you have to explain the bridge to the present state on observational evidence, yet another chicken egg puzzle that cannot be waved away with a just so story. KFkairosfocus
February 27, 2015
February
02
Feb
27
27
2015
05:41 AM
5
05
41
AM
PDT
Joe:
Talk about being naive. If a 2 part structure does one thing and you need another function tat the two part structure doesn’t have, then what?
Who said anything about 'need'? If you have a 2 part structure, another part can be added to it. If that triumvirate is beneficial, or at least minimally detrimental, you've taken a step. If it isn't, you stay where you are.Hangonasec
February 27, 2015
February
02
Feb
27
27
2015
05:37 AM
5
05
37
AM
PDT
KF - nonetheless, the incremental argument needs to be addressed, not wafted away. Given that incremental interaction, mutual tuning and loss of prior supporting structures has the potential to generate apparent IC, we would need to eliminate that possibility. The heart of the system is the 'motor', a rotary ATPase. It has the capacity to generate weak directional motion alone, which would certainly be beneficial against stationary competitors, but would soon be eliminated by swifter, better-steered ones. Attachment of spindle fibres would enhance the efficiency of coupling, and so those weaker movers would soon be outcompeted, and that step of evolutionary history erased. Nearly all the proteins in the flagellar assembly have homology with other cellular proteins, so virtually nothing new is evidenced. I know that designers bodge as well, but the point is that little fundamental evolutionary novelty is displayed by the flagellar system, rendering it at least potentially within reach of evolutionary tinkering and tuning. Knocking out one of the proteins now says little about its optionality when first incorporated.Hangonasec
February 27, 2015
February
02
Feb
27
27
2015
05:33 AM
5
05
33
AM
PDT
F/N: To get an idea, consider the main gear of the 6500 C3 reel (and the gear train and drag). Just the geat. Set up an exact coord system, and acurately locate and align a brass disk with circularity, and accurately alighned sides. Hard already and with multiple IC requisites. Then, define a gear tooth profile and cut it along a precise trajectory along the rim. Increment to the next. Continue and have a precise integer number of teeth around the disk. HALT at just the right number. Cut a precisely centred keyway and a precisely machined cylindrical cut-out for the drag disk stack. Notice, this is already well beyond 125 bytes to describe much less to execute. Real world IC usually implies FSCO/I. Proceed in a similar manner for the rest of the parts, then assemble correctly for function. IC is a reality and a commonplace. In the world of life, things like the flagellum are notorious, but so are co-ordinated communication and control networks, and of course in that context the von Neumann kinematic namomolecular self replicator integrated with a metabolising, encapsulated automaton with smart gating. IC is relevant, starting with OOL. Go on to any number of systems in complex body plans, of which bird flight is perhaps the clearest and most familiar. You may want to quibble on borders and definitions, but the reality is there, staring you in the face. KF PS: Don't overlook the implications of the transcendental number nature of pi, a gear can be precise but not exact. Something has to give in circumference or diameter as the two are not commensurate.kairosfocus
February 27, 2015
February
02
Feb
27
27
2015
05:25 AM
5
05
25
AM
PDT
Hangonasec:
If a 2 part structure can evolve, any additional number of parts can be added serially by the same means.
Talk about being naive. If a 2 part structure does one thing and you need another function tat the two part structure doesn't have, then what?Joe
February 27, 2015
February
02
Feb
27
27
2015
05:21 AM
5
05
21
AM
PDT
Piotr @ 124- what is your point? If you have a 5 component structure that requires all 5 components before it achieves the function what do you do? And if all the parts have to be specific, ie can't be just any part, what do you do?Joe
February 27, 2015
February
02
Feb
27
27
2015
05:20 AM
5
05
20
AM
PDT
H & P: The issue is, that multiple specifically matched core parts need to be correctly arranged and coupled for function in a context where if one is missing, no function. Thus incremental achievement is blocked. And, exaptation runs into Menuge's C1 - 5:
For a working [bacterial] flagellum to be built by exaptation, the five following conditions would all have to be met: C1: Availability. Among the parts available for recruitment to form the flagellum, there would need to be ones capable of performing the highly specialized tasks of paddle, rotor, and motor, even though all of these items serve some other function or no function. C2: Synchronization. The availability of these parts would have to be synchronized so that at some point, either individually or in combination, they are all available at the same time. C3: Localization. The selected parts must all be made available at the same ‘construction site,’ perhaps not simultaneously but certainly at the time they are needed. C4: Coordination. The parts must be coordinated in just the right way: even if all of the parts of a flagellum are available at the right time, it is clear that the majority of ways of assembling them will be non-functional or irrelevant. C5: Interface compatibility. The parts must be mutually compatible, that is, ‘well-matched’ and capable of properly ‘interacting’: even if a paddle, rotor, and motor are put together in the right order, they also need to interface correctly. ( Agents Under Fire: Materialism and the Rationality of Science, pgs. 104-105 (Rowman & Littlefield, 2004).]
That makes blind arrangement and coupling -- especially in an automaton that is self-assembling -- highly unlikely to the point of effectively impossible, especially when one blends in population genetics and available timeline. KFkairosfocus
February 27, 2015
February
02
Feb
27
27
2015
05:07 AM
5
05
07
AM
PDT
DEF'N, adapting Behe slightly:
What type of biological system could not be formed by “numerous successive, slight modifications?” Well, for starters, a system that is irreducibly complex. By irreducibly complex I mean a single system composed of several well-matched interacting parts that contribute to the basic function, wherein the removal of any one of the [core] parts causes the system to effectively cease functioning. [DBB, 1996, p. 39, emphases and parenthesis added.]
--> I would have preferred, multiple and specifically arranged and coupled . . .kairosfocus
February 27, 2015
February
02
Feb
27
27
2015
05:01 AM
5
05
01
AM
PDT
If a 2 part structure can evolve, any additional number of parts can be added serially by the same means. Once A+B has arisen in such a co-dependent manner, AB + C can arise, then ABC + D, and so on. And parts can be lost. Evolution is a bugger for erasing history.Hangonasec
February 27, 2015
February
02
Feb
27
27
2015
05:00 AM
5
05
00
AM
PDT
1 4 5 6 7 8 11

Leave a Reply