"Devolution" and gene loss in evolution

Not really, when you consider the number of generations and years of passage involved.

You got me. That statement would seem to deny neutral theory. But correct me if I am wrong–the fixation of most mutations is due to drift, but there is still room for selection to act on favorable alleles. It’s not all drift.

First of all, your objection that the failure to take the two step path was due to some error seems a stretch to me. We checked the plasmids, we checked the ability to revert singly, we checked for the effect of plasmid copy number, etc. Our conclusion is couched the way thousands of other papers are, when we say “May nonetheless be unattainable.” Did we make a general conclusion? No. would other tests be valuable ? Yes. It would be appropriate to ask how much over-expression is too much, and to do it in another system. All worth doing.

I’m sorry. This doesn’t make sense to me. Showing that cells reduce expression of unneeded genes is not a failure.

Here is the full context of what you quoted:
This scenario is relevant to the origin of new genes. Most
genes are thought to have originated by a process of gene duplication
and divergent evolution leading to new function [7-10].
Typically, experimental studies of this process look for genes
whose products are able to metabolize a new compound or replace
a missing function [11-13]. But since the recruited gene
product often performs its new function very poorly, it is likely
to require over-expression to have selective benefit.

Emerging functions, at least enzymatic ones, are typically weak. Scientists who study recruitment of proteins to new functions typically do so by mutating the gene and expressing it from high copy high expression plasmids. They can often detect functional change precisely because so much protein is being made. In nature, however, such overexpression is unlikely, and so such recruitment would be unlikely to occur in nature. I can think of one case by Copley’s group McLoughlin, S. Y. and S. D. Copley. 2008. A compromise required by gene sharing enables survival: Implications for evolution of new gene activities. Proc. Nat. Acad. Sci. USA 105:13497-13502. where they had one protein serve two functions, made possible by modest overexpression.

Your paper is a case where the insertion of an IS element provided new promoter activity, activating the expression of the enzyme necessary to metabolize the new carbon source. I am not denying that such things occur. However, they are not the same as recruitment of an existing gene to a new function by modification of that gene’s sequence.

@T_aquaticus, thanks for the excellent paraphrasing!

@agauger,

If we look at T’s sentence that I have highlighted above, note that it is essentially analagous to what happened with the common ancestor of snakes.

The snake’s 4 limbed method of traveling was not the best match for the ecosystem or niche that this reptile found itself. In terms of phenotype, the snake had two paths ahead (if he was going to remain in that ecosystem):

  1. Develop a brand new kind of legs (perhaps jointed differently), or

  2. jettison all four legs.

Perhaps it was just a “bad day” or a “good day” … but that population of reptiles headed down the road towards eliminating all 4 legs.

Just to demonstrate that this wasn’t inevitable, there are reptiles of the snake family that actually still have legs. The link below is for a snake known only in fossil form. YEC publications have attempted to use this fossil as a way of proving the Genesis story about snakes having legs.

But this is a flawed assessment. If this snake was extinguished by the flood, then what it really shows is that God put a curse on snakes… and missed a bunch!

http://science.sciencemag.org/content/349/6246/416

That is just trading one term for another. A gene includes its promoter, so the ability to express the cit gene in aerobic conditions is a new gene function.[quote=“agauger, post:18, topic:36902”]
The ID people I think you are referring to, who will not credit a gain of information even from the foundation of the vertebrate lineage (slight hyperbole) are probably of the genic kind. And extreme. But is a rearrangement a gain of information? In the citrate case it would definitely be a gain of function mutation IMHO. But new genetic information? Don’t know.
[/quote]

If you know of any evolutionary changes that they classify as the emergence of new genetic information then I would be interest in reading about it.

Parasitism can also induce some of these same changes. We see some parasites losing their digestive tract or other adaptations they would need if they were living out on their own (e.g. tapeworm). Some of the smallest and simplest genomes known are found in microorganisms called obligate intracellular parasites that live inside of other cells. In fact, viruses may have once been fully fledged microorganisms of their own (still hotly debated if memory serves), but over time their genomes were stripped down to the point that they couldn’t reproduce on their own anymore.

Paring down a body plan or genome is evolution in the same way as adapting new body parts or gaining new genes. Evolution doesn’t have a set goal other than an increase in fitness.

2 Likes

I have a ton of first hand knowledge when it comes to the laboratory and molecular biology that spans the last 20 years. I remember the “hard way” of doing things, and still love to use capillary transfer and paper towels for blotting DNA. :wink: Funding is difficult to get in the best of times, so I empathize with your position. Here’s to hoping that you get the lab you deserve!

What I am curious about is if there is any carbon source or “food” left once the culture reaches 1E8/ml. Do wild type or trp+ bacteria reach a higher density in the same media (MDMC)?

Also, don’t feel that is necessary to answer these questions if you are limited on time.

ORFan genes.

I couldn’t immediately find the answer to your question as posed–it may be in lab notebooks. But figure 3 of the paper does compare growth of cells with plasmids in MDMC (minimal medium supplemented with salts and glucose (.2% I believe) and tryptophan) and with MDMC plus 1% glucose. The growth appears to be essentially the same.

If I am reading the title of the figure correctly it is talking about growth on solid media instead of liquid media. There are 1-2 generation differences in those results, but I don’t know if that would carry over to liquid culture. There is also the possibility of bacteria using up resources without increasing CFUs due to limits on translation.

Of course, no experiment is perfect and people can nit pick them to death. Like I said, I don’t expect all of my questions to be answered so no need to spend more time than you think appropriate.

In an article you wrote for Evolution News & Science Today, you wrote:

“In other organisms, similar stretches of DNA are non-functional, whereas in our genomes they are transcribed and, in many cases, translated into protein. They lack similarity in sequence to other coding genes anywhere in the catalog of all known protein-coding genes. Because they have no relatives in other genomes, they are called orphan genes. As such, they present a strong challenge to the Darwinian story. This is because in Darwinian terms, repurposing genes requires gaining some means of turning on transcription and translation and acquiring a useful function.”
reference.

I guess I don’t see why ORFan genes pose a problem for the theory. As far as I can tell, Darwinian evolution is not limited to already existing genes. If a mutation produces a new promoter region and results in the transcription and translation of DNA downstream of that new promoter I don’t see why this would be a problem for the theory. I have looked at a few human specific ORFan genes, and orthologous DNA can be found in the chimp genome for those human specific ORFan genes, so it isn’t as if a whole new sequence has to emerge in the human lineage.

I was also curious if there were any examples of evolution producing “new information” that ID proponents agree with. It seems that you are arguing against evolution producing ORFan genes, or perhaps I am misreading the article you wrote.

I’m sorry, I misread your question. I thought you were just asking for new information, not new information produced by evolution.

Orphan genes are a really hot area of new research. There are several things required to get an protein-coding orphan gene from non-coding sequence. It’s more than just a promoter. First the stretch of DNA has to be free of stop codons, which is statistically unlikely for an ORF of any length in the absence of selection. (Most orphans tend to be shorter, which is evidence in favor of an evolutionary origin. Second a promoter and regulatory elements, terminator sequences, and most importantly, a stable functional protein needs to be translated from the ORF. There are groups trying to establish that random sequences can produce functional protein. All this needs to be established as possible before claiming NS and RM are responsible for orphans.

As for examples where i think evolution has produced new function, therefore new information, there is the case of colobine monkeys developing new digestive enzymes in response to a change in diet–it was gene duplication and recruitment to a new function of an existing enzyme as I recall.
https://bip.weizmann.ac.il/education/course/evogen/evogen2003/Duplication1/zhang_ng_2002.pdf.
http://www.sciencedirect.com/science/article/pii/S0168952502027555
Some of the examples in Dennis’s book are trivial examples. Anything with a step-wise selectable path leading to new function qualifies. Nylonase would be an example.

I know that many internet bloggers will deny an increase in information unless it’s the appearance of novel sequence, a la orphans. But Doug Axe and I draw the line for what evolution can’t do at genuinely new innovation, on the order of a new enzymatic activity with no shared substrate or reaction chemistry. If all it takes are a few point mutations that confer a benefit, then sure, the “new” function might be considered a gain in information.

Added edit: Then again it might not be an increase in information if all that happened is the enhancement of a minor activity already present. We get into semantics here. Nylonase is an enhancement of a probably pre-existing activity–not clear-- but it in any case it produced a new function–digesting nylon. And it was a simple evolutionary path. Two selectable steps. Gain of information? Depends. New genetic material? No. New function? Yes.

That matches what I have read, especially Ruiz-Orera et al. (2015):

“Using comparative genomics, we show that the expression of these transcripts [ORFans] is associated with the gain of regulatory motifs upstream of the transcription start site (TSS) and of U1 snRNP sites downstream of the TSS.”

A straightforward study could compare the orthologous regions of the human and chimp genomes for human ORFans to see how many differences there are. I would be interested in seeing how an ID proponent would determine which of those changes random mutagenesis is allowed to make and which can not be produced by those known mechanisms.[quote=“agauger, post:33, topic:36902”]
As for examples where i think evolution has produced new function, therefore new information, there is the case of colobine monkeys developing new digestive enzymes in response to a change in diet–it was gene duplication and recruitment to a new function of an existing enzyme as I recall.
https://bip.weizmann.ac.il/education/course/evogen/evogen2003/Duplication1/zhang_ng_2002.pdf.
http://www.sciencedirect.com/science/article/pii/S0168952502027555
Some of the examples in Dennis’s book are trivial examples. Anything with a step-wise selectable path leading to new function qualifies. Nylonase would be an example.
[/quote]

Thanks for the example. It helps when you can see a few positive examples instead of all negative ones.

I would break the problem into pieces. First, can you get functional proteins from random sequences? A recent paper has claimed so, but there are reasons to doubt the work. More work will be needed.
Second, how difficult it it to assemble a promoter in front of a potentially functional sequence? We have some hints from bacterial work that it can be done by insertion of an IS element or rearrangement of an existing promoter. Does that work in eukaryotes or does it have to happen one base at a time?
Third temination sequences. Same questions.
Fourth regulatory control.
Fifth, is there a selective advantage at each step? Can they be acquired one at a time or must two or more be required together?

We know the answer to that: yes, by definition. Given that functional proteins arise from polypeptide sequences, any random peptide library has a nonzero probability of containing functional sequences, and a sufficiently large library is essentially guaranteed to contain functional sequences. A more valid question is this: what is the probability that a particular random polypeptide will have a function (in a given context)? The one thing we know about this probability is that it is greater than zero.

Which paper are you referring to? There are several that I know of. One very recent paper is this one:

https://www.nature.com/articles/s41559-017-0127

What are the reasons to doubt this work? Again, we know that random sequences can contain functional peptides; that’s not even in question.

1 Like

You are no doubt aware that there have been studies of exactly what you were asking here. These values range from one in 10 to the 63rd power or one in ten to 74th power, depending on how you want to frame the statement. Now that’s a very rare set of proteins to have to go searching for in a library. That’s why it’s important for us to examine the more recent attempts at generating random libraries of protein sequences and testing them for function. We have conflicting data that need to be reconciled, and I submitted that there may be unforeseen complications in the Neem et al paper which is the one I think you’re referring to.

Are you talking about estimates for the existence/formation of folds? Because that’s not the same question. Those numbers sound like the ones proposed by Doug Axe, and you are no doubt aware that the paper is not widely cited and is not influencing the ongoing exploration of the protein universe. If you know of other studies of protein function in sequence space, I’d love to discuss them with you.[quote=“agauger, post:37, topic:36902”]
I submitted that there may be unforeseen complications in the Neem et al paper which is the one I think you’re referring to
[/quote]

I haven’t seen your comments on Neme et al. but would love to read them. In fact there have been a few recent papers on this topic, and there is the classic from Keefe and Szostak below.
https://www.nature.com/nature/journal/v410/n6829/full/410715a0.html

2 Likes

That would be a rather open ended experiment because there are literally millions of possible functions that would need to be tested for. Competition assays and measures of fitness are one way to go about it, but that would also screen out deleterious functions. Afterall, function is function.

Also, ORFan genes may not be truly random sequence to begin with. You would have to pay careful attention to GC content, homologies to transposons, and other considerations.

Since the chimp and human genomes are almost fully sequenced this should be pretty straightforward. Look for the sequence differences between genomes with reference to chimp and human specific ORFan genes.[quote=“agauger, post:35, topic:36902”]
Fifth, is there a selective advantage at each step? Can they be acquired one at a time or must two or more be required together?
[/quote]

The first step may even be slightly deleterious, but with enough background mutation you are bound to produce promoters sequences in intergenic regions. If the initial product is slightly deleterious there can be later mutations that increase the fitness of the new gene.

Those studies looked at just one function out of millions, if they are the studies I am thinking of. You can’t screen for just B-lactamase activity, as an example, and claim that a sequence has no function if it fails to cleave B-lactams. There could also be very divergent sequences that produce very different tertiary structures that have the same activity which means that tweaking just one protein fold in one protein isn’t going to tell you the full range of possible proteins with that function.