Dr. Stephen Meyer and Dr. Douglas Axe were recently interviewed by author and radio host Frank Turek on the significance of November’s Royal Society Meeting on evolution, in London. The two Intelligent Design advocates discussed what they see as the top five problems for evolutionary theory:
(i) gaps in the fossil record (in particular, the Cambrian explosion);
(ii) the lack of a naturalistic explanation for the origin of biological information;
(iii) the necessity of early mutations during embryonic development (which are invariably either defective or lethal) in order to generate new animal body types;
(iv) the existence of non-DNA epigenetic information controlling development (which means that you can’t evolve new animal body plans simply by mutating DNA); and
(v) the universal design intuition that we all share: functional coherence makes accidental invention fantastically improbable and hence physically impossible.
In today’s post, I’d like to focus on the third argument, which I consider to be the best of the bunch. The others are far less compelling.
Over at the Sandwalk blog, Professor Larry Moran and his readers have done a pretty good job of rebutting most of these arguments, in their comments on Professor Moran’s recent post, The dynamic duo tell us about five problems with evolution (January 14, 2017). Larry Moran’s earlier 2015 post, Molecular evidence supports the evolution of the major animal phyla cites a paper by Mario dos Reis et al. in Current Biology (Volume 25, Issue 22, p2939–2950, 16 November 2015) titled, “Uncertainty in the Timing of Origin of Animals and the Limits of Precision in Molecular Timescales,” which convincingly rebuts Meyer and Axe’s first argument, by showing that animals probably originated in the Cryogenian period (720 to 635 million years ago) and diversified into various phyla during the Ediacaran period (635 to 542 million years ago), before the Cambrian. I might add that we now have strong evidence that anatomical and genetic evolution occurred five times faster during the early Cambrian, at least for arthropods – although as Intelligent Design advocates have pointed out, that still leaves unanswered the question of how animal body plans arose in the first place.
Meyer and Axe’s second argument asserts that natural processes are incapable (as far as we can tell) of creating significant quantities of biological information – and especially, new functions or new anatomical features. Much of the argument rests on the alleged rarity of functional proteins in amino acid sequence space – a claim that was crushingly refuted in Rumraket’s recent post on The Skeptical Zone titled, Axe, EN&W and protein sequence space (again, again, again) (October 12, 2016). As for the claim that natural processes can’t create new functions, it’s simply bogus. The following three papers should be sufficient to demonstrate its empirical falsity: Five classic examples of gene evolution by Michael Page (New Scientist Daily News, March 24, 2009), Evolution of colour vision in vertebrates by James K. Bowmaker (Eye (1998) 12, 541-547), and Adaptive evolution of complex innovations through stepwise metabolic niche expansion by Balazs Szappanos et al (Nature Communications 7, article number 11607 (2016), doi:10.1038/ncomms11607).
I’m not really qualified to discuss Meyer and Axe’s fourth argument, but it seems to me that Professor Larry Moran has addressed it more than adequately in his recent post, What the Heck is Epigenetics? (Sandwalk, January 7, 2017). The last four paragraphs are worth quoting (emphases mine):
The Dean and Maggert definition [of epigenetics] focuses attention on modification of DNA (e.g. methylation) and modification of histones (chromatin) that are passed from one cell to two daughter cells. That’s where the action is in terms of the debate over the importance of epigenetics.
Methylation is trivial. Following semi-conservative DNA replication the new DNA strand will be hemi-methylated because the old strand will still have a methyl group but the newly synthesized strand will not. Hemi-methylated sites are the substrates for methylases so the site will be rapidly converted to a fully methylated site. This phenomenon was fully characterized almost 40 years ago [Restriction, Modification, and Epigenetics]. There’s no mystery about the inheritance of DNA modifications and no threat to evolutionary theory.
Histone modifications are never inherited through sperm because the chromatin is restructured during spermatogenesis. Modifications that are present in the oocyte can be passed down to the egg cell because some of the histones remain bound to DNA and pass from cell to cell during mitosis/meiosis. The only difference between this and inheritance of lac repressors is that the histones remain bound to the DNA at specific sites while the repressor molecules are released during DNA replication and re-bind to the lac operator in the daughter cells [Repression of the lac Operon].
Some people think this overthrows modern evolutionary theory.
So much for epigenetics, then.
The fifth and final argument discussed by Drs. Meyer and Axe relates to the universal design intuition. I’ve already amply covered both the merits and the mathematical and scientific flaws in Dr. Axe’s book, Undeniable, in my comprehensive review, so I won’t repeat myself here.
The “early embryo” argument, helpfully summarized by Dr. Paul Nelson
That leaves us with the third argument. Looking through the comments on Professor Moran’s latest post, it seems that very few readers bothered to address this argument. The only notable exception was lutesuite, who pointed out that examples of non-lethal mutation in regulatory DNA sequences are discussed in a paper titled, Functional analysis of eve stripe 2 enhancer evolution in Drosophila: rules governing conservation and change by M.Z. Ludwig et al. (Development 1998 125: 949-958). The paper looks interesting, but it’s clearly written for a specialist audience, and I don’t feel qualified to comment on it.
As it turns out, I wrote about the “early embryo” argument in a 2012 post, when it was being put forward by Dr. Paul Nelson. Nelson handily summarized the argument in a comment he made over at Professor Jerry Coyne’s Website, Why Evolution Is True:
Mutations that disrupt body plan formation are inevitably deleterious. (There’s only one class of exceptions; see below.) This is the main signal emerging from over 100 years of mutagenesis in Drosophila.
Text from one of my Saddleback slides:
1. Animal body plans are built in each generation by a stepwise process, from the fertilized egg to the many cells of the adult. The earliest stages in this process determine what follows.
2. Thus, to change — that is, to evolve — any body plan, mutations expressed early in development must occur, be viable, and be stably transmitted to offspring.
3. But such early-acting mutations of global effect are those least likely to be tolerated by the embryo.
Losses of structures are the only exception to this otherwise universal generalization about animal development and evolution. Many species will tolerate phenotypic losses if their local (environmental) circumstances are favorable. Hence island or cave fauna often lose (for instance) wings or eyes.
Obviously, loss of function is incapable of explaining the origin of new, viable body plans for animals.
A hole in the argument?
On the face of it, Nelson’s three-step argument certainly looks like a knock-down argument, assuming that the premises are factually true. But are they? A commenter named Born Right made the following response to Dr. Nelson over at Jerry Coyne’s Website (emphases mine):
Paul Nelson,
Lethal mutations will kill the embryo. But what you’re totally failing to understand is that not all mutations are lethal. Many are tolerated. I heard you cite the example of HOX gene mutations in flies and how altering them kills the embryos. You didn’t mention the entire story there. Do you know that there are wild populations of flies having HOX gene mutations? Even in the lab, you can create viable HOX-mutant flies that have, for example, two sets of wings. In fact, simple non-lethal mutations in HOX genes can profoundly alter the morphology. It is these non-lethal mutations that natural selection “cherry picks”, provided they confer a survival advantage on the organism.
Many mutations actually arise as recessive mutations, not as dominant ones. They spread through the population remaining dormant or having a mild effect, until there is a sufficient number of heterozygotes. Then, interbreeding between heterozygotes will cause homozygous mutations to arise suddenly throughout the population. If the new feature improves survival & reproductive success, it gets rapidly selected…
Macroevolution is a gradual response to climate change and other environmental pressures. Organisms accumulate non-lethal mutations that changes their body plan bit by bit until they are well adapted to their changing habitat.
However, a 2010 Evolution News and Views post co-authored by Dr. Paul Nelson, Dr. Stephen Meyer, Dr. Rick Sternberg and Dr. Jonathan Wells, contests the claim that Hox gene mutations are non-lethal. The authors assert that such mutations are, at the very least, defective:
Mutations to “genetic switches” involved in body plan formation … disrupt the normal development of animals. With the possible exception of the loss of structures (not a promising avenue for novelty-building evolution, in any case), these mutations either destroy the embryo in which they occur or render it gravely unfit as an adult. What the mutations do not provide are “many different variations in body plans.”…
… [T]here are solid empirical grounds for arguing that changes in DNA alone cannot produce new organs or body plans. A technique called “saturation mutagenesis”1,2 has been used to produce every possible developmental mutation in fruit flies (Drosophila melanogaster),3,4,5 roundworms (Caenorhabditis elegans),6,7 and zebrafish (Danio rerio),8,9,10 and the same technique is now being applied to mice (Mus musculus).11,12
None of the evidence from these and numerous other studies of developmental mutations supports the neo-Darwinian dogma that DNA mutations can lead to new organs or body plans–because none of the observed developmental mutations benefit the organism.
Indeed, the evidence justifies only one conclusion, which Wells summarized in his last slide at SMU:
“We can modify the DNA of a fruit fly embryo in any way we want, and there are only three possible outcomes:
A normal fruit fly;
A defective fruit fly; or
A dead fruit fly.”
The Wikipedia article on Drosophila embryogenesis may interest some readers.
What I would like to know is: are the Hox mutations in fruitflies mentioned by Born Right in his comment above neutral or deleterious – and if the latter, are they only slightly deleterious or highly deleterious?
A follow-up comment by Born Right
In a subsequent comment over at Why Evolution Is True, Born Right cited two scientific references in support of his claims:
Paul Nelson,
Fantastic new research shows how fish developed limbs and moved onto land. Boosting the expression of Hoxd13a gene in zebrafish transforms their fins into limb-like structures that develop more cartilage tissue and less fin tissue!
http://www.sciencedaily.com/releases/2012/12/121210124521.htm
http://www.sciencedirect.com/science/article/pii/S1534580712004789
Importantly, the overexpression of Hoxd13a in zebrafish was driven by a mouse-specific enhancer. This shows that the regulatory elements acting on the enhancer are present in both fishes and distantly-related mammals!
The first paper, titled, From fish to human: Research reveals how fins became legs (Science Daily, December 10, 2012) is written in a style that laypeople can readily understand. I’ll quote a brief excerpt (emphases mine):
In order to understand how fins may have evolved into limbs, researchers led by Dr. Gómez-Skarmeta and his colleague Dr. Fernando Casares at the same institute introduced extra Hoxd13, a gene known to play a role in distinguishing body parts, at the tip of a zebrafish embryo’s fin. Surprisingly, this led to the generation of new cartilage tissue and the reduction of fin tissue — changes that strikingly recapitulate key aspects of land-animal limb development. The researchers wondered whether novel Hoxd13 control elements may have increased Hoxd13 gene expression in the past to cause similar effects during limb evolution. They turned to a DNA control element that is known to regulate the activation of Hoxd13 in mouse embryonic limbs and that is absent in fish.
“We found that in the zebrafish, the mouse Hoxd13 control element was capable of driving gene expression in the distal fin rudiment. This result indicates that molecular machinery capable of activating this control element was also present in the last common ancestor of finned and legged animals and is proven by its remnants in zebrafish,” says Dr. Casares.
This sounds fascinating, and to me it constitutes powerful evidence for common ancestry, but the real question we need to address is; exactly how early in the course of the zebrafish’s embryonic development did these mutations take effect?
The second paper cited by Born Right (“Hoxd13 Contribution to the Evolution of Vertebrate Appendages” by Renata Freitas et al. in Developmental Cell, Volume 23, Issue 6, pp. 1219–1229, 11 December 2012) is much meatier, because it’s the original papaer on which the Science Daily report was based. The authors contend that “modulation of 5′ Hoxd transcription, through the addition of novel enhancer elements to its regulatory machinery, was a key evolutionary mechanism for the distal elaboration of vertebrate appendages,” and they conclude:
Within the developmental constraints imposed by a highly derived teleost fin, our results suggest that modulation of Hoxd13 results in downstream developmental changes expected to have happened during fin evolution. This, together with the evidence we provide that the upstream regulators of CsC were also present prior to tetrapod radiation, makes us favor an evolutionary scenario in which gain of extra 5′ Hoxd enhancers might have allowed the developmental changes necessary for the elaboration of distal bones in fishes that evolved, ultimately, into the tetrapod hand.
This sounds a lot more promising, but after having a look at it, I’m still rather unclear about exactly how early these hypothesized mutations would have had to have occurred, in the course of vertebrate embryonic development. Perhaps some reader can enlighten me.
Well, that’s about as far as my digging and delving has taken me. I’d like to throw the discussion open at this point. Are there any known examples of early embryonic mutations which are not deleterious, and do they shed any light on how new animal body plans might have evolved? Over to you.
(Note: the image at the top [courtesy of Wikipedia] shows the ventral view of repeating denticle bands on the cuticle of a 22-hour-old Drosophila embryo. The head is on the left.)
I’m claiming that there are no genes without homologes, that must have appeared out of the blue, as if they were engineered and put there for a purpose, and I’m arguing that if there were such genes and they were involved in early embryological development, they would have a larger impact on development than perhaps a single AA substitution, and that gene insertion would be either far more deleterious (if we accept that disrupting development at early stages is problematic) or produce a considerably different creature, ergo, saltation. I think it’s common sense
So what? Cars traveling down the road is routinely observed.
Actually cancer is due to random effects on the design
And they just happen is the smarter position- really?
Laws of physics No pixies pushing required.
Do you really think that items which are engineered just appear out of the blue?
Are there YouTube videos? I don’t believe anything I can’t see on YouTube.
So? Such a finding would in no way create a disturbance in the force.
There’s no reason a new gene cannot arise in one species which would have no homolog in any other species. In fact, that’s how evolution is supposed to work.
Ok I got it-
Pixies pushing Pixels (say it 5 times fast)
I don’t. In fact, I believe that is exactly what is happening. That explains why all the atoms don’t just fall to the ground.
WTF? can you follow the conversation here, or are you going to pull a dazz?
WTF?
Allan:
Mung:
Gee, I wonder how one might find out?
OK, you were presumably referring to genes arising from previously non-functional sequences. I don’t think saying “that’s how evolution is supposed to work” is accurate since there are many other sources of variation, and it’s completely irrelevant to the topic at hand anyway
I’ve argued in this discussion that characteristics of the cell are not totally owed to the DNA, but to non-DNA mechanism of inheritance. The problem in recognizing non-DNA inheritance is the fact there aren’t single points of failure in non-DNA inheritance mechanism in contrast to DNA which is very much single point failure. If one mutates a nucleotide in the genome, one expects that the next generation of DNA (barring back mutations) will have that mutation.
In contrast, with things like organelle inheritance, a single failure or change in an organelle (like the mitochondrion) will not necessarily matriculate to the next generation because there are so many other organelles in the same class. An example of this was pointed to me by genetic engineer Rob Carter:
http://isogg.org/wiki/Heteroplasmy
But that is the DNA, not the proteome, and the way the proteome connects to each other to form a mitochondrion is what is inherited, not just the DNA.
There is a wide variety of mitochondrion. From that stellar outfit, Cold Spring Harbor:
So how do these mitochondira evolve since the inheritance of the mitochondrial structure, as I’ve shown, is not purely DNA? It must be the entire mitochondria itself, and not only that, the “mutation” (not just DNA) must be global to all mitochondria in the cell.
Thus, this is proof DNA does not contain all the heritable information. In the case of organelles like mitochondria, it is outside the DNA. I prefer not to use the term epigenetic to describe this, but rather organelle inheritance.
So is there research on organelle inheritance? A little, but it doesn’t fit the gene-centric narrative. Nevertheless here the best summary from an old paper that was pioneering in the field:
http://www.cell.com/cell/abstract/S0092-8674(00)81284-2
But this is amazing:
Another organelle that can be structurally inherited is the centriole. The mother centriole serves as a physical template for the daughter centriole. Furthermore, if the mother centriole is damaged, a de novo centriole can be synthesized, so there are two mechanisms of creating the same organelle! One by structural inheritance, and the other by a set of instructions residing elsewhere in the cell. Amazing!
What I’m trying to say is that cells are 3D printers of themselves, or somewhat like 3D photocopiers. Part of the 3D blue print is the existing cell itself and parts in the cell. It’s not just inside the DNA!
And heteroplasmy is also quite rare. This is because of population genetics happening within the cell: drift and selection happen to the mitochondrial population in each cell, and mitochondria undergo a severe bottleneck in every organismal generation too.
…the great majority of which are encoded by the nuclear genome, not the mitochondrial genome. Did you know that?
Mitochondria, being former bacteria, evolve in more or less the way that bacteria evolve, except that most of their evolution happens in the nuclear genome. Whatever does this have to do with epigenetic inheritance? You understand that the proteins and other molecules in those mitochondria are products of genomes, right? What in that organelle is inherited over many generations other than the genome itself?
What you describe is just not true in any non-trivial sense.
newton,
Thornton’s work is trying to find an ancestral track for enzymes. Catalyzing small molecule reactions.
dazz,
Big claims take big balls. Hats off to you Dazz 🙂
Obviously Bill and Mung are far too dumb to follow my argument, here’s hoping Vincent can provide some valuable insight..
Why don’t you just show him wrong by citing a few genes that don’t have homologs? Should be easy for you.
Declaring homology is begging the question
How can we test that claim beyond saying mitochondria look like stripped down versions of some bacteria?
Agree with many qualifications, and the ID community needs to reframe their argument to make it better.
Almost any random object of appropriate size can be used as a paper weight. It hardly means complex function can emerge easily, it only means relatively simple function can emerge simply provided there is already an infrastructure that can use it. This is like saying a randomly shaped key is automatically functional because we can find, or at least put together a lock that will match it. Or even easier, a random alphabetic string is automatically functional because we can make it a functional password since we can create a login account somewhere that will use it.
Rumraket accurately saw flaws in the way ID proponents define functional proteins. There are signaling proteins, like insulin, a so-called peptide hormone. It is functional in many vertebrates, but plants and bacteria have no use for it. What this means is we could hypothetically concoct a random polypeptide, and with provision of the proper infrastructure, probably make it functional just like we can take a random alphabetic string and make it functional by letting it be password or some sort of encryption key or whatever.
So Rumraket crushed a weakly stated ID argument. I will agree to that extent. Well done for him, but that doesn’t mean the IDists can’t improve their fundamental claim that such systems are highly improbable.
Going back to the example of insulin. We have transgenic creatures with human insulin in them for research and manufacture for diabetes patients. Scientists have put reformatted insulin genes in bacteria and yeast. The bacteria and yeast have no use for insulin. So let’s say hypothetically before the emergence of vertebrates some random mutation created an insulin molecule in a bacterium or yeast. Do we declare it functional because hypothetically some yet-to-exist life form like a vertebrate could use it even though the bacterium where the insulin emerged has no use for it?
What makes insulin functional is not the protein itself, but the entire system that makes it functional from the Receptor Tyrosine Kinase where the insulin molecule docks to all the machinery in the cell that processes the insulin signal. Also a means of insulin synthesis and regulation are needed like Beta cells etc. We know if insulin is misregulated it is lethal.
So I agree, Kudos to Rumraket for crushing a bad ID argument, but it doesn’t crush the improbability of a complex functional system emerging randomly.
No, it didn’t. In the Axe experiment, the requirement was to “fold into the beta-lactamase domain and break down ampicillin”. That’s much, much more specific, and doesn’t in any way tell you about the relative frequency of “proteins that bind other proteins and catalyze breaking them down”.
By the way, ampicillin is not a protein. There was nothing in Axe’s experiment about proteins binding other proteins.
Once again you’re just blathering about stuff you don’t know anything about.
stcordova,
Mitochondria have DNA. Although people talk of them as ‘cytoplasmic’ – eg Cytoplasmic Male Sterility, which is down to mitochondria ‘trying’ to avoid the sperm dead end – These Are Not The Exceptions You Seek.
I don’t know why Creationists ride this “It’s not DNA” horse so furiously. I can only assume it boils down to a dislike for any view associated, however loosely, with Richard Dawkins.
But when it leads people to deny even the basic biology that has been elucidated over the last 50-70 years (DNA was not even definitively accepted as the hereditary material c1950), and make ignorant statements such as “there is no*** evidence DNA controls development”, one can only gawp.
There may be subtleties emerging, but I have yet to see an exception to the current paradigm – ie a system that has no root in DNA.
*** That’s right, “NO evidence”!.
Mung,
No point mutations, indels, frameshifts, duplications or subgene recombinations then? No translation of previously untranslated sequence? Just ‘protein sequence’ out of the blue, knitted by pixies.
No, but that brings up an important point I’ve thought about myself. Which is that function is context-specific. Which means the ID argument from “density of functional molecules in sequence space” swings right back in their face.
For any given molecule we take from life, mutate and test it, there’s some chance if we change the conditions of the test (temperature, pH, pressure and so on), we can resurrect a function it initially appears we broke. But we can’t know that without testing it. How many different combinations of conditions are possible? It seems to be there’s an infinite number of combinations.
Environment and local conditions determine what is or is not a functional molecule. There isn’t some “true” number to practically get at (it might exist in theory, but the combinatiorial space is too large for us to be able to estimate it with limited experiments).
Is your protein sensitive to temperature fluctuations and prone to denature? Then lower the local temperature. Add stabilizing metals or minerals. Change the salt-concentration or pH.
Concerning how many possible biomolecules are possibly functional in some environment, you can almost always change the surrounding conditions to “make it functional again”. Change one thing so it appeas to no longer function, you can resurrect it by changing the other components to compensate.
Suppose we pick some enzyme from E coli and splice it into a hyperthermophilic archaea living inside a hydrothermal vent. Say, the now famous “strain 121”. There’s a pretty good chance NONE of the molecules from E coli will function in strain 121 at in it’s natural living conditions. Strain 121 can live and grow at 121 degrees celcius. It can survive, but is dormant at 130 degrees C. Likely all the proteins expressed by the E coli genome will misfold, denature and probably hydrolyse rather quickly in strain 121.
37 degrees C with pH 7 is one among an infinity of possible combinations. Anything that works there will fail in an archaea living in the superheated walls of an acidic black smoker. From the perspective of Strain121, we look like impossible beings. None of our biomolecules will work where it lives.
Were we the ones living in acid at 121 degrees C, changing the sequences of our proteins into the sequences that work in E coli, would break them from our perspective. We would put those mutations in there and test them at the conditions where we live, and see every one of them fail. And then we’d mistakenly conclude that “this cytochrome C sequence no longer functions”. But it does function, at 37 C in E coli.
Take another example: Binding of transcription factors to specific DNA sequences. So there’s some sequence of DNA that is a target for a regulatory protein. We mutate this DNA enough, and the regulatory protein can no longer bind that piece of DNA. Have we now proven that the DNA, and/or the protein is truly “non-functional”? No, of course not, because strictly speaking just as we can mutate the DNA-sequence, so can the regulatory protein also be mutated and keep being able to bind the sequence. It is context-specific. You quickly realize there’s no possible sequence of DNA that could not be bound by some protein molecule. Technically all possible DNA-sequences could work as a transcription factor binding site. The only important thing is that the quence is “different enough” from other sequences in the genome, that the protein won’t bind the wrong area accidentally.
stcordova,
Most mitochondrial proteins come from the nucleus. Those that do are like any other protein that comes from the nucleus. They root in DNA. Those that don’t have a different inheritance pattern. They too root in DNA. So, taken as a whole …
That’s why I’m begging you to reconsider. You can learn the essentials rather quickly if you unlearn the way Larry is trying to teach his faithful vs. the way professional epigenetic/epigenomic institutions (like the NIH ENCODE and NIH RoadmapEpigneomics consortiums) actually teach epigenetics.
Larry and UncommonDescent’s focus on acquired epigenetics traits is cringe worthy. It conflates Larmakian and Lysenkoist ideas about acquired traits with the real important role of epigenetics (the far better term is “chromatin modification”) in development and embryogenesis.
Larry Moran as quoted by VJTorley:
Methylation is trivial. Following semi-conservative DNA replication the new DNA strand will be hemi-methylated because the old strand will still have a methyl group but the newly synthesized strand will not. Hemi-methylated sites are the substrates for methylases so the site will be rapidly converted to a fully methylated site. This phenomenon was fully characterized almost 40 years ago [Restriction, Modification, and Epigenetics]. There’s no mystery about the inheritance of DNA modifications and no threat to evolutionary theory.
I have to keep pouncing on this because this is distortion! Methylations of DNA are not solely a consequence of the semi-conservative DNA replication process! In fact a lot of methylations and demethylations happen on DNA outside of the mechanism Larry describes. Larry gives the impression this is the only way DNA cytosines are methylated or de methylated. Not true.
I showed this diagram that refutes Larry’s characterization:
http://www.nature.com/nature/journal/v403/n6769/fig_tab/403501b0_F1.html#figure-title
earlier from this paper:
http://www.nature.com/nature/journal/v403/n6769/full/403501b0.html
Changes in the methylation patterns are very important to making a cell a stem cell or not. We’ve been able to induce adult cells to be come pluripotent stem cells by changing the methylation patterns, and hence we have an avenue to no longer need embryonic stem cells from aborted fetuses.
Methylation patterns are different between cell types. Those change in methylation patterns does not follow the pathway Larry describes because Larry is describing only one way methylation patterns are put, and insinuates that’s the only way. The other ways are far more complex, they are not well-understood, but we know they are important for development.
If Larry’s reporting had been a little more balanced, he would highlight studies like this (which would actually refute his points, so it’s obvious he won’t):
https://genomebiology.biomedcentral.com/articles/10.1186/gb-2013-14-8-131
How these levels of epigenetic complexity evolved is the real problem. How is it that brain cells have different methylation patterns than other cells in the body? For that matter, why and how do so many cells have different methylation patterns than other cells in the body?
Larry focuses on a situation of copy-and-paste methylation in the germ line, he totally ignores emergence of different methylation patterns in different cell types in the somatic lines. Where is the re-programming information stored that makes these differing patterns. No one knows!
Below is a picture of the methylation patterns in a normal (CTRL) rat’s DNA and that of an epileptic (EPI) rat’s DNA. The issues is how the right methylation patterns evolved in the first place, and how these differing patterns end up in different cells.
Compare the methylation pictures of CTRL (normal rat) to that of EPI (epileptic rat).
It is from this essay:
Also, as pointed out earlier. Humans have different methylation patterns than chimps, even though their DNA is highly similar.
So, again, don’t be too quick to provide blessings to Larry’s characterization of epigenetics. It’s idiosyncratic and doesn’t agree with the characterization of institutions that actually research epigenetics as attested to by publications on the matter. Contrast Larry’s claim “Methylation is trivial” and dismissal of histone modifications with actual published research that deals with DNA methylation and histone modifications:
http://www.roadmapepigenomics.org/publications/
Agree, but that’s not the point I’m making. The mitochondria provides provides the structural template for the proteins coded by the genes. The nuclear DNA and mtDNA do not provide blueprints for the final structure (the assembled mitochondrion), only blue prints for the parts (proteins).
The only poor analogy I can think of is that of a photocopier. Ink and paper are the raw substrates to make a copy. There maybe separate recipes and process for making ink and paper, but the blue print for an image must come from a pre-existing image, not from the recipe or process for synthesizing inks and making paper.
The original image provides a template for assembling the ink onto the paper to make a copy. Nuclear DNA and mtDNA provides recipes for the proteins that serve as the raw materials for the mitochondria, but a pre-existing actual mitochondrion provides the 3D “image” that serves as a template to make other mitochondrion.
Structural inheritance is analogous to the 3D photocopying process which made a 3D photocopy of a wrench that Leno is holding (below):
http://www.popularmechanics.co.za/wheels/rapid-prototyping/
But the 3D copying in life is far more intricate. We don’t know how it is done. We do know the DNA provides instructions to make the raw materials, but I argue emphatically it is unlikely DNA is the sole source of the developmental/assembly instructions. We have strong experimental hints to that effect such as with the mitochondrion and cilia of paramecium.
This is also apparently the case for other organelles like the Golgi, Endoplasmic Rectilium, centrioles, chloroplasts (in plants), etc.
If mitochondria serve as structural templates for future generations of mitochondria, how did all the varieties of mitochondrial organelles evolve in eukaryotes except by “mutating” the actual physical structure (in addition to mutating the DNA).
We were able to induce structural “mutations” in cilia that persisted. See:
http://dev.biologists.org/content/develop/105/3/447.full.pdf
The paper also uses the term “phenocopies”. I like that.
Some of the assembly of organelle’s strikes me more like 3D photocopying of phenocopies, not purely gene expression. Variations of this 3D photocopying process can then be used to create different cell types. How this is done, no one knows yet.
Thornton’s work involves ancestral sequence reconstruction for protein families from life, regardless of whether they’re enzymes and “catalyze small molecule reactions” or not. And to be clear, they’re often part of larger complex functions and machines.
In one of their experiments, they reconstructed part of the evolution of the increased complexity of the ATP synthase molecular motor. They actually showed how a number of components were added to the machine, what mutations casused this, what the ancestral and descendant structures were like and how they functioned.
Evolution of increased complexity in a molecular machine
Gregory C. Finnigan, Victor Hanson-Smith, Tom H. Stevens & Joseph W. Thornton
Holy crap you spew a lot of misinformation sal. One doesn’t have the time to take this shit apart day in and day out. Are you employed to misinform?
OMG, NO ONE KNOWS. So GOD must have done it, and evolution must be false. And there’s just no way methyltransferase activity are themselves regulated by a classic environmentally sensitive regulator.
You read it here first!
Must it? Actually no. It is entirely possible to generate an image, or a complex 3-dimensional structure, from a process, or simple program, rather than from a “pre-existing” image or structure.
It’s what the evidence says. Deal with it
Wrong, Allan. The argument is that DNA does NOT determine what will develop. DNA controls and influences development it just doesn’t determine what type of organism will develop.
Influencing and controlling is not the sane as determining. I know you don’t care what the actual genetic researches have to say but please read the following:
stcordova,
I really don’t see how nuclear proteins for internal structural elements are any different from any others. Sure, they accrete to a scaffold. Same goes for new lipid molecules. But they all root in DNA. The arrangement depends in part upon what is already there, but that does not knock DNA off its perch.
Nobody ever proposed a naked DNA molecule repeatedly clothing itself in other molecules from scratch. That’s not what the centrality of DNA is about.
stcordova,
Because when you are talking about the relevance of epigenetic inheritance, that’s the kind of thing one ignores. The inheritance of epigenetic mechanism (genetically mediated, AFAIK), is not the same as the inheritance of epigenetic state.
I’ll hazard a guess: it is under genetic control. What a wild, out-there-speculatin’ kind of guy I am!
Rumraket,
A bit like this
I see nothing!!
Infinite complexity, and indeed beauty and harmony, in so simple a rule.
He has also reconstructed ancient proteins per Answers in Genesis.
Rumraket,
Thanks for the paper. I look forward to reviewing this. I do understand that ampicillin is not a protein. My understanding was that beta-lactamase is made up of two proteins sub units and what Axe modified was one of the sub units. If this is not the case I apologize.
newton,
Rumraket showed me a paper attempting to show a reconstruction of ATP phase and how a few mutations can improve function. I stand corrected.
Bu…But… but… the sequence space!
We may know some day where some of it is when we can build the high throughput methods to sequence the proteome, the glyclome, the glycoproteom, the epi-transcriptome, the glycol lipidome, epi-proteome (all the postranslation enhancemens in proteins other than histones, which we still can’t do that well anyway).
When we see how staggering the complexity of the cell truly is (vs. the 82.3 megabyte picture Larry Moran and Dan Graur are peddling), then it be a little more obvious God did it. Praise be.
The proteins in a living frog before he is put in a blender and spun around are about the same as after he’s put in a blender and spun around.
I think you’re missing the essential point that it’s HOW the proteins are connected to each other, not how the proteins are made. Did you not even read how changes in the phenocopy of the cilia are transgenerationally inherited?
And we don’t even have any handle of the state of the glycol-protein conjugates, free floating glycans, the glycol-lipid conjugates, the post-translational enhancements on the individual amino acids on the proteins themselves. We don’t have the high throughput methods yet to determine these things.
We may find out otherwise, especially since we can duplicate enucleated cells that just leverage mRNAs or other ncRNAs
If we can somehow pump the requisite proteins into a organelles as they duplicate so they don’t need those mRNAs, then we’ll know it’s not ultimately the genes. Even Warren (the pioneer of organelle inheritance), pointed out the creation of in-vitro Golgi stacks. That experiment would seem a good one to look at since presumably he had a Golgi that was decoupled from the nucleus.
Agree, but I was talking about photocopiers. 🙂
Yes, he has admitted to abusing children for money with his mischaracterizations of science.
He’d obviously like a more lucrative role in the intelligent design creationism movement.
Nice projection…
stcordova,
Yep, the well-known phenomenon of the ‘you-never-know-ome’.
That’s right. It does not need a DNA molecule to be present perpetually. But … where do those mRNAs and lncRNAs come from?
I mean, why can’t you join these dots yourself?
stcordova,
I had a biochemical pathways chart pinned to my wall at uni. It was rilly complex. And that was ages ago. I am unimpressed, and unsurprised, by the additional complexity since discovered. Far too complex to be designed, if we are simply trading opinions.
I do think design advocates need to demonstrate that design is possible, by which I mean, demonstrate that foresightful modification is possible.
How does one know the effects of a modification on the ecosystem?