DNA is not just a static read-only memory (ROM) for coding proteins, but hosts dynamic random access memory (RAM) in the form of methylations and histone modifications for regulation of gene expression, cellular differentiation, learning and cognition, and who knows what else. The picture below depicts how rapidly the RAM aspect of DNA is changed during embryogenesis.
Many of the DNA methylation patterns are in non-coding repetitive regions. This suggests at least some of the roles of non-coding DNA are involved in supporting the complex epignomic memory in each cell.
Depicted below are changes in epigenetic methylation marks on the DNA in the stages of embryo development. The light green colors indicate epigenetic methylations and the darker blue colors indicate absence of epigenetic methylations. In boxes “a” through “l”, the bottom part is the DNA from the mother and the top part is the DNA from the father. Eventually the DNA from mom and dad mix in the 4 cells of box “m”.
Note how the epigenetic marks are erased from the parternal DNA first!
The depiction below shows how rapidly epigenetic changes happen even in time frames as short as hours. Each cell has a slightly different methylation pattern and hence each cell’s RAM has some unique information. If we consider that the human has 100 trillion cells and that each cell has 30 million potential methylation sites, the sum total of RAM memory implemented by epigenetic cytosine methylation alone is on the order of sextillions of bits of Shannon information. Like histones, DNA methylations can be written, erased and read.
When scientists inhibit epigenetic changes, the results are usually lethal. So we know the epigenetic component of the DNA is vital to life.
a–e, Anti-5-methylcytosine (MeC) immunofluorescence of mouse one-cell embryos. a, Zygote 3 h after fertilization with intense MeC labelling of both pronuclei (>10). Numbers in parentheses indicate the number of embryos analysed. b Paternal and maternal pronuclei at 6 h (>10). c, Undermethylated paternal pronucleus at 8 h (>20). The smaller female pronucleus remains methylated. d, Aphidicolin-treated one-cell embryo displaying demethylation of the male pronucleus (>20). e, First metaphase (>5). f–j, Controls. Anti-DNA immunofluorescence of one-cell embryos demonstrates that both chromatin sets are accessible to antibody molecules. f, 3 h (>5). g, 6 h (>5). h, 8 h (>5). i, Aphidicolin treatment (>5). j, First metaphase (2). k,l, MeC staining of two-cell embryos at 22 h (>20) (k) and 32 h (>20) (l) shows that the paternal and maternal compartments have different methylation levels. m, Four-cell embryo at 45 h (>10). The MeC-staining intensity of the maternal half of the nucleus is weaker than in two-cell embryos. Scale bar, 10 mum.
http://www.nature.com/nature/journal/v403/n6769/fig_tab/403501b0_F1.html#figure-title
http://www.nature.com/nature/journal/v403/n6769/fig_tab/403501b0_ft.html
What does your red herring have to do with the NIH Roadmap/the journal Nature definition of epigenetics which I’ve used? The OP was about the methylation marks on DNA in the early stages of development. I called those epigenetic marks. Do you have problem with that or do you still insist methylation marks are peripheral to epigenetics?
TomMeuller made the following claim here:
“Memory propagation is done by transcription factors.”
That’s an abusrd statement, imho. You agree with that DNA_jock? How about you Allan Miller?
Here is the definition of a transcription factor:
https://en.wikipedia.org/wiki/Transcription_factor
Oh, I understand how this game is played. When I make an assertion, you’ll demand a citation. When I actually provide a citation, you’ll accuse me of cutting and pasting and not understanding the citation even though it proves my point and disproves my detractor’s criticism. I just pointing this out, because such whining really belies the fact I scored a point on matters of fact.
In fact transcription factors are blocked the configuration of histone modifcations.
Therefore histone modifications inhibit or facilitate transcription factors from binding. If histones modifications can regulate transcription factor binding, then it is absurd to say as Tom did, “Memory propagation is done by transcription factors.“.
I’ve stated, it is not well known where ALL the epigenetic memory resides. We know where some of it may reside and is copied from, but not all.
Tom did get one thing right:
Yes, dynamic Random Access Memory. 🙂
For interested reader’s benefit, the enzymes called DMNT’s (de nove methyl transferases) are involved in some of the copying of epigenetic marks in somatic cells. This diagram is from the prestigious scientific journal Nature.
The right side of the diagram illustrates how DNA is replicated where one methylated DNA double strand is forked into two identically methylated DNA double strands. So when DNA is duplicated, it’s DNA methylation marks (which are epigenetic marks) can be duplicated as well.
This is significant in trying to understand the photo of embryos in the OP. A different mechanism is in play in the copying of methylation marks in the zygote than the way DNMT does it in other somatic cells because clearly the marks in the zygote are erased only to be put back after being erased!
In my humble opinion this look just like a reboot of the marks in the zygote, wiping the slate clean and starting with a rebooted clean set of necessary marks. Where and how the information for these epigenetic marks is stored and inherited transgenerationally is not understood.
We know some of the mechanism of how these marks can be passed down in somatic cell lines — it’s through mechanisms like DMNT. How it is done in the Zygote after the marks are wiped clean, no one really knows. Additionally, the methylation marks are not exactly copied between each cell since each cell type has different methylation marks as well as different methylation marks depending on the phase of the cell.
http://www.nature.com/nrg/journal/v2/n1/images/nrg0101_021a_i2.gif
Sez Sal:
Well, that would scarcely be an ethical experiment. OTOH, stuff Gaur or Yak or Zebu DNA into a cow cell and you get…what exactly?
Your humble opinion is wrong. We know that transcription factors do provide memory propagation. And we know how.
We also know that DNA methylation can provide memory propagation, and we know how.
We have some ideas about how histone modifications affect replication timing, and how replication timing affects histone modifications, but I do not believe there is much in the way of support for the idea that histone modifications are “propagated” in the way that DNA methylation can be, but is not always.
Your idea that the specific modifications are somehow ‘stored’ elsewhere and are re-written following erasure is hilarious. What evidence do you have to support this goofy idea?
Your argument
Makes me laugh. By the same token, red blood cells are inherited: look – I have them, my grandmother had them, see! They must be inherited. No, Sal, it’s the DNA that codes for them that is inherited.
Think
Still waiting to hear about mules and hinnies…
But not all memory propagation, certainly not enough to propagate the memory in histone modifications since the histone modifications can block the binding of transcription factors.
Or really, so when the paternal DNA in the picture in the OP Block 3, has the methylation marks erased, where does the information come from to put the methylation marks on the paternal DNA? Given we have imprinting specific to either maternal or paternal DNA, it can’t be coming from the maternal side, and it can’t be coming from the paternal side (since it was just erased).
You’re just laughing at your own illogic.
DNA_Jock,
Same animals. Just because we call them different species doesn’t make it so.
Also red blood cells are not cells but formed elements. However the egg membrane from mom, along with its microtubule array are also inherited, along with the DNA
How does unguided evolution explain transcription factors?
Thanks for affirming the OP — methylation marks as Random Access Memory.
So are the epigenetic marks on imprinted genes inherited or not. Allan Miller insists they aren’t inherited. What say you? 🙂
You agree then with Allan Miller’s claim that conserved epigenetic marks on imprinted genes aren’t inherited? 🙂
How about you TomMeuller,
Do you agree with Allan Miller and DNA_jock that conserved epigenetic marks on imprinted genes are not heritable? 🙂
stcordova,
Let me try again. When a set of chromosomes is placed in a sperm, they have a particular methylation state A (characteristic of the male). When a set of chromosomes is placed in an egg, they have a particular different methylation state B (characteristic of the female).
When a sperm fertilises an egg, it finds itself first in almost exclusively ‘female’ cytoplasm. Before the chromosome sets are even in the same nucleus, the methylation states are wiped – first the male’s, then the female’s. Therefore, the bulk of methylation states in the embryo are clearly NOT inherited, by any reasonable definition of inheritance. How can they be if one methylation pattern comes from Mom and one from Dad, and both are wiped?
Nonetheless, something is not wiped. A small region remains differentially methylated after the wiping process – the DMR. So, the vast majority of methylation marks are certainly NOT inherited by the offspring, but a tiny nucleus of difference remains. It appears to be there that the ‘memory’ of whether the chromosome came from Mother or Father is held – in the DMR, on chromosomes.
From that DMR nucleus, remethylation spreads out, in a pattern that is different according to the origin of the chromosomes, triggered in part by the DMRs.
But at the next generation, even the DMRs are wiped. The DMR of the entire chromosome set is established according to the gender of the individual generating the gametes, not either parent. It does not depend on any methylation state in any prior generation – ie, the methylation state is not inherited, only the capacity to generate a particular state from a particular set of conditions (which is pretty much the same as any other gene action).
So, I don’t see any way you can really rescue your notion that imprinted marks are inherited across multiple generations, other than persisting in a strange, frankly obtuse, conception of inheritance. They are established differentially in male and female gametes, which drives remethylation following fertilisation. The methylation marks you provide in your sperm are not, in any way, inherited from your dad or your mum. They are set during gametogenesis according to your gender. It is a 1-shot flag.
That’s pretty amazing. Thanks for posting.
Is this DMR region present in single celled organisms and/or haploid organisms? Is it present in plants?
The whole point is to ‘inform’ the next-generation diploid of the parental origin of individual chromosomes. It makes zero sense for this information to be carried further. One does not need information re: the gender of the grandparent (there is, in any case, only room for half of them to be retained).
Mung,
Imprinting in general is restricted to mammals ‘above’ monotremes.
It appears to be associated with internal development of offspring (though is absent in viviparous reptiles).
The best supported theory for its origin is actually an evolutionary one. Nearly all imprinted genes have foetal expression, and the imprinting tends to favour resource utilisation (the ‘resource’ being the mother) that is optimal for the imprint, not the organism. The result is an approximate no-win situation, as you’d expect given that genes spend half their lives in each gender.
It makes little adaptive (or Design) sense to have parental genes in a foetus tussling over optimal resource usage. But as a result of evolutionary conflicts of interest between male and female, the effective Mexican standoff that we see makes sense.
Obtuse? There are the epigenetic marks that are unique to paternal DNA on imprinted genes on parent and child and grand child and great grand child — otherwise this often leads to disease (as listed in the links I provided).
Did those marks appear by accident or were they inherited? If they were inherited, it refutes your assertion that paternal epigenetic marks on imprinted genes aren’t inherited.
Here is study that showed paternal imprinting.
http://bmcgenet.biomedcentral.com/articles/10.1186/1471-2156-5-13
How did that paternal imprint get there? Voodoo coincidence or inheritance from the father or mother.
If you say the paternal imprint came from the father, then where was the information stored to make the imprint on the paternal gene? If you say I came from the mother where was the information stored to make that imprint? In both cases, it has to be in the cytoplasm if the imprints were erased as depicted in block C of the picture in the OP. The sperm has a small amount cytoplasm as can be seen
here
What would be amazing is if both cytoplasms contribute to the imprinting. No one knows where the information for paternal imprinting comes from after the erasure event depicted in block C of the picture in the OP.
If one argues the paternal imprint is NOT one of the regions wiped during the erasure process, then that also is transgenerational epigenetic inheritance. That is also a possibility, and in which case mechanism like DMNT1 are sufficient to perpetuate the imprints.
Bottom line, in every case, the epigenetic marks on the paternal DNA of an imprinted gene are inherited. Your assertion is indefensible that epigenetic marks on imprinted genes are not inherited.
Note, this is a different question as to what degree of acquired epigenetic marks during the life of an individual become transgenerationally heritable. We know many do not become transgenerationally heritable for the simple fact many of them are in somatic cells, and that is probably a good thing.
There is a growing field of non-genetic inheritance. I’ve believe a lot of heritable information is not stored in the DNA coding sequences, but in a variety of ways. DNA carries necessary but not sufficient heritable information.
The discussion of paternal epigenetic marks on imprinted genes shows that some information must be transmitted outside of the coding sequences of DNA.
http://www.nature.com/npp/journal/v40/n1/full/npp2014127a.html
stcordova,
Yes, of course there are! I can’t believe you think I don’t know this. So, think it possible that you might be misunderstanding me? The same genes are paternally imprinted on paternally sourced chromosomes in generation after generation after generation. Ditto female.That’s how imprinting works, FFS! But the marks aren’t inherited. They are replaced by the marks appropriate to the current individual’s gender, on gametogenesis.
The mechanism to place the imprint on a chromosome repeatably in a particular environment is inherited – because, of course, all (epi)genetics roots in DNA, like I have been saying. The imprint is placed by genes, influenced by DMRs. This does not mean that an epigenetic mark is copied from one generation to the next. It is established, according to gender, in generation after generation.
False dichotomy. It is environmentally conditioned. In the environment of a male chromosome, one kind of mark will be established. In the environment of a female chromosome, another kind of mark will be established. This is not inheritance of the mark.
Can you really not see the difference between copying a methylation state and re-establishing it?
I’m sure you’ll get great marks in your exams. But on this showing, you won’t learn anything. Obtuse is the word. I wonder if your apparent weakness on inheritance may explain your apparent inability to think – even hypothetically – in evolutionary terms.
stcordova,
You do realise that that is ONE GENERATION, don’t you? That is precisely how long an imprint needs to last, and also the maximum that a given gender imprint can last.
How would the information about an individual chromosome’s gender at G0 survive passage through multiple generations, passing through both male and female bodies? Why would it need to? You only need to know the gender now to set the gamete imprints. The genome doesn’t need to (and can’t) know whether a given autosome or X was in a male or a female >1 generation back.
stcordova,
There is no evidence that every last bit of information is erased. Imprinted genes occur in blocks, and methylation spreads across regions. It only requires a diagnostic DMR to remain after the primary erasure to allow remethylation from that DMR outwards to be done differentially.
But, before you start, that residual DMR itself will be reset on gametogenesis in the next generation.
Sal – you have still managed to fail the Turing Test!
Let’s recap, I threw you out as many life lines as I could without giving you the answer:
Let’s give Sal another hint…
Sal – your answer must include three words: Hinny, Mule & some version of the word “erase”.
Here is another hint: Your task will be made easier if you actually refer your own recent post and to the very diagram you yourself posted – at 12:12 pm when you commented that “a picture is worth a thousand words”.
That would be your post right here:
Now I find it most intriguing that you yourself posted this very link from
http://learn.genetics.utah.edu/content/epigenetics/imprinting/
… and that you yourself still don’t get it – Allan needs to patiently – most patiently explain and reexplain (and without success I may add) what you yourself posted.
Interesting very interesting indeed.
Well, the three keys words were hinny, mule & some version of the word erase.
Sal, you clearly cannot even remember his own posts! Earlier you even wrote wrote:
I asked the professor where the imprinting information is stored during the erasure process in the first 3 hours after fertilization and he indicated know one knows!!!! It is a topic of future research.
Exactly what Allan was trying to get you to wrap your head around. (BTW – you prof’s answer was not entirely accurate).
This much is clear: Sal’s original incarnation of Epigenetics as RAM is blatantly wrong. His newest rescue attempts clearly contradict his original thesis (without his realization no less) and still have not answered my original question.
ITMT – Sal has demonstrated blatant lack of understanding and insight
Time’s up. This exchange is no longer amusing and I am pulling the plug.
Hi Allan
And your DMRs explain the hinny/mule conundrum how exactly?
I am not clear how DMRs can simultaneously remain “diagnostic” and be “reset”
ITMT – I really must be getting old and losing my touch… exactly what is the distinction between DMR & facultative heterochromatin?
Thanks in advance for helping an aging Biology teacher keep current.
best
But the paternal imprint still appears again in the next generation and the next and the next — so much so it’s been declared as “conserved”. So what if it disappears briefly (like for a few hours in box C). This is just like the blackboard analogy I gave earlier. We have a parent black board and a child blackboard with identical or almost identical markings. The child black board gets erased and then the erased markings get re-written to match the parent — at least at imprinted sites. Does that still not count as copying the parents marking? In the case of paternal imprints is that an accident or was that inherited?
So was that imprint inherited or is it just a voodoo coincidence? How about you DNA_jock or Tom Meuller. Are conserved paternal epigenetic marks that appear generation after generation inherited or not? 🙄
So Tom Meuller, do you insist DNA methylation marks are peripheral to epigenetic inheritance (either germ or somatic cell lines)? Take note of the de novo methyl transferase diagram I provided earlier. Looks to me the epigenetic marks on the parent strand provides a template for the epigenetic marks on the daughter strand in a DMNT1 activity scenario.
A simple yes or no will do. Do you insist DNA methylation marks are peripheral to epigenetic inheritance (either germ or somatic cell lines)?
So I take your word over a researcher of imprinted genes? No dice.
If the a female zygote gets a paternal imprint, it has to be reset to the maternal version, but where does the knowledge of the correct imprint come from? It must be inherited.
If one wants to classify this as not evidence of inheritance, to me, that’s just mincing words because clearly the knowledge and technology to make those imprints does not come from scratch, it is inherited.
Allan has a different notion of inheritance than my usage. I think my usage is more accurate.
http://www.nature.com/scitable/topicpage/Genetic-Imprinting-and-X-Inactivation-1066
Btw, RNAs are not usually referred to as transcription factors as Tom Meuller was trying to advocate:
Transcription factors are generally viewed as PROTEINS not RNAs. If you want to use another convention, go ahead, but let’s not pretend it is the one mostly in use.
https://en.wikipedia.org/wiki/Transcription_factor
stcordova,
Yes, because everyone has a father, so everyone receives male-imprinted chromosomes.But if the individual is a girl, the paternal imprints are thrown away at F1, and re-marked as female. Even if the individual is a boy, the paternal imprints are thrown away and re-marked as male (as are the maternal chromosomes). The imprints are not copied – they indicate origin-of-parent; you don’t need to pass them more than one generation along. This is not an example of inheritance of an epigenetic mark, it is an example of inheritance of a gene that causes an epigenetic mark.
You misunderstand what ‘conserved’ means in that context. You need to pay particular attention during the evolution modules, because this is where your evolutionary blind spot really becomes apparent. When a gene displays evolutionary conservation, it is the underlying sequence, not its methylation state, that is being referred to. Being methylated appears to have a conservative effect on the underlying sequence – possibly because only certain base combinations are capable of methylation, possibly because evolutionary stability prevents oppositely imprinted genes from diverging as readily as less constrained cases.
Then it has not been inherited. Something else has been inherited – the thing that writes the marks. Every diploid genome has the capacity to mark both male and female imprints. It does both during embryogenesis but only one (binary male or female) during gametogenesis. The latter marks replace the former.
No.
1) You have a bunch of sheets of paper, You write ‘M’ and ‘Y’ on half, and ‘M’ and ‘X’ on half. Your wife writes writes ‘F’ and ‘X’ on all of hers.
2) You each take a sheet and put them together.
3) Someone rubs out the M and F, though not completely
4) The M and F are inked in again.
time passes …
5) The Ms and Fs are rubbed out and
a) If the papers both contain X’s, an ‘F’ is written on both papers.
b) If the papers contain an X and a Y, an ‘M’ is written on both.
In neither a) nor b) is the previous mark on the papers copied. It is set from the value of the X and Y. It may or may not be the same two generations running. It cannot be the same indefinitely.
Obviously:
Step 1 is gametogenesis in you and your wife.
Step 2 is fertilisation.
Step 3 is pronucleus erasure
Step 4 is re-establishment
Step 5 is gametogenesis in the child.
If it is 5a, the M goes no further.
If it is 5b, the F goes no further.
Even if you see two generations of M as ‘copying’, how many generations does it take to truncate this sequence?
stcordova,
I have a fantasy where one of your profs takes a sock here and tries to teach you something. My expectation is that you would disbelieve his every word. Just ‘cos.
stcordova,
This is completely unrelated to the multi-generational inheritance of imprinted methylation marks. Do you agree, yes or no?
stcordova,
I have never disputed (and have explicitly insisted) that the information required to produce a methylated imprint is inherited. But all along you have been linking lengthy and irrelevant articles about the direct copying of a methylation state from chromosome to chromosome. As far as parental imprinting is concerned, across multiple generations that does not happen. Do you agree?
If I had access to a researcher I’d be inclined to ask Allan what question I could ask that researcher to better understand the point that Allan is trying to make. I’d be more likely to learn something, right or wrong. Declaring victory benefits nothing.
By the way Allan. I bought a biology book. So in a few days I’ll know more about the subject than you. Just sayin’.
Wait for it.
Tom Mueller cites Mark Ptachne:
Ptachne doesn’t even mention this paper that refutes his point:
TomMueller,
Differentially Methylated Regions result in differential gene expression according to the parent of origin of the chromosome. About half of methylations turn genes on and half turn them off. Imprints tend to operate in the foetus, and male genes are selected to take more from the mother; female genes are selected to oppose this.
So my reading of the mule/hinny case is that imprinted genes often end up acting in antagonistic pairs, but with close evolutionary history, as would be expected in a breeding pool, these pairs are at a standoff and the net payoff is equalised between male- and female- oriented imprints. For more distant relations, however, there are more asymmetries when these distant cousin genomes find themselves together. These asymmetries would tend to result in some hybrid differences that depend simply on which species was the father and which the mother.
This would accord with Haig’s kin selection theory.
I’d like to see a Creationist explanation for imprinting…
The reset need not be complete. Imprinted genes occur in blocks. Erasure involves removal of most meythlation, but if sufficient remains (perhaps in non-transcribed regions), it can still allow mechanistic distinction by the re-methylation process. The latter tends to spread laterally in both directions from an already-methylated core.
Note that methylation must occur twice – once faithful to parent of origin during embryogenesis, and once based solely upon gender of individual during gametogenesis. In the latter case, the ‘wipe’ must be complete; in the former, something must (logically) remain.
The former is a methyl mark on the DNA bases themselves, the latter a more exotic range of rearrangements on the protein, I guess.
Mung,
Now, to determine whether to take this statement at face value or not … [presses ‘detect’ button, whirrr, click, hiss]
At least we can agree on something. The imprint is conserved, thus the information required to produce this conserved imprint is inherited.
No, but that is because we have different notions of what constitutes inheritance. Your definition of inheritance suggests that for an imprint to be inherited that the parent imprint has to serve as template for the child imprint much like DNA serves as template. You want to use that definition, that’s fine, it’s not mine, and I’ll tell you why.
There seems to be a lot of epigenetic heredity that is not reducible to simple “copy the template” approach as found with DNA replication. The copying of histone modifications during cell replication is not trivial and is not a copy-the-template type of copying, but it is still called heritable between somatic cells as in this article:
http://www.nature.com/nrm/journal/v10/n3/full/nrm2640.html
Anyway, thanks for persisting and for the conversation. I hope that at least clarifies why we don’t agree, we have different notions as to what constitutes non-genetic inheritance.
stcordova,
It’s got nothing to do with our notions of inheritance. There is a specific biochemical process A that copies a methylation state from one strand to another. There is a different biochemical process B that establishes a methylation state de novo.
I am saying (and have been saying all along) that process A is not involved in multigenerational inheritance of imprinting – that imprints are not copied from strand to strand. Do you now agree? Because all your references to date, supposedly refuting my contentions, have been about Process A.
stcordova,
Be that as it may, the genes that generate and wipe imprints do reside on DNA.
Sal,
To say that some state is “heritable’ through mitosis in no way implies that it is heritable through generations. Think beta-globin expression levels…
Allan has explained to you very clearly and cogently that it is the apparatus that imprints, and not the imprinted pattern itself, that is heritable through multiple generations, and you still fail to get it. I am tempted to suggest that you print this thread, and ask your profs to comment on it. But they already have quite the cross to bear, so I won’t…
Oh, and to answer your goofy repeated question, I still agree with Allan 100%.
Agreed, I posted at least twice on DMNTs.
I said so particularly for the Zygote, but for also the differences in methylation that is obviously in evidence in somatic cell lines.
That may be true of paternal imprints. I have not said process A was involved in that, in fact I said quite the opposite here:
So far you can’t say I said something in opposition to what you said.
I never said they were, I said the opposite so do insinuate that I said one thing when I actually said the opposite.
That is not true in light of the fact that I said a different process that DMNT copying was in play.
But clearly some sort of information for the pattern of marks must be stored and inherited somewhere, and I specifically said it was not being copied from the paternal DNA as I pointed out to DNA_jock here:
You are arguing against positions I don’t hold nor made in this this discussion. You successfully refuted arguments I never made. I have to credit you for that.
Heck, Sal knows more biology than you after taking just a few classes. I bought an entire book!
Perhaps DNA_Jock will share his EA with you.
😉
Baloney!
Allan has clearly explained a position I don’t hold and he refuted a position didn’t articulate. You’re just repeating the same mischaracterization.
Baloney. I said as much when I pointed out your laughing at your own illogic here:
Do you have issues with reading comprehension? Oh, maybe you were confused because I said “Block 3” which is technically “Block C” even though Block C is the third block in the photo.
So what’s your excuse for attributing arguments to me which I didn’t make. By the way, if the pattern is appears generation after generation it is heritable, it doesn’t mean however that it is copied from the actual imprints itself any more than the skeleton of child is copied directly from the skeleton of the parents.
I never said the imprints were copied from pre-existing imprints, I said the opposite especially since I’m the one who posted the photo of the paternal DNA getting its methylation marks wiped (as depicted in Block C of the OP).
So, you’re the one who fails to get it.
Wow – amazing really… Sal still has not figured out his new version of events contradicts what he originally wrote.
It’s late and I need to hit the sack so I shall toss in a couple of mischievous hand-grenades and take my leave.
Allan – Could you please provide me references for these residual islands of (cis-acting allele discrimination) methylation besides centromeric and pericentric repeats?
And as my last random act of mischief – I just want to mention the Överkalix study, Marcus Pembrey and colleagues observed that the paternal (but not maternal) grandsons of Swedish men who were exposed during preadolescence to famine in the 19th century were less likely to die of cardiovascular disease. If food was plentiful, then diabetes mortality in the grandchildren increased, suggesting that this was a transgenerational epigenetic inheritance. The opposite effect was observed for females—the paternal (but not maternal) granddaughters of women who experienced famine while in the womb (and therefore while their eggs were being formed) lived shorter lives on average.
Take home message – epigenetics memory can fade over several generations, not just one.
Second take home message, Human Genetics seem to mirror equine and murine epigenetics as detailed above in a most peculiar fashion. Ocham’s razor suggests to me that trans-acting factors seem the best explanation and if I understand correctly, trans-acting allelic discriminating factors have been determined if not yet fully characterized.
stcordova,
(cross-posted from Noyau)
But the structure isn’t inherited. The epigenetic marking goes via DNA inheritance, same as everything else.
There is a very real and non-trivial distinction at work here. Obviously the developmental processing that makes an organism a member of that species is stably inherited, including the involvement of epigenetic silencing and activation. The doctor doesn’t hand over an undifferentiated bucket of slop to the happy parents – “Congratulations Mrs Goldstein, it’s a boy!”
That is genetic inheritance of epigenetic behaviour.
But the process you are trying to big up is epigenetic inheritance of epigenetic behaviour. The imprinting case is a model example where THE LATTER DOES NOT HAPPEN. It can’t, for reasons I have painstakingly laid out.
These are real mechanistic distinctions. Because if you want some sense in which genetics is bypassed (perhaps you hope that if Mendel is wrong then so is Darwin), you have to look somewhere other than imprinting . Because imprinting is genetically inherited.
That you accuse me of mincing words to save a position is rich.
stcordova,
And you did so in supposed refutation of my contention that imprinted epigenetic marks do not survive replication across more than 1 organismal generation. ie, it was irrelevant.
And you did so in supposed refutation of my contention that imprinted epigenetic marks do not survive as such for more than 1 organismal generation. ie, it was irrelevant.
Whatever you may have said to someone else, This is what you said to me at the start:
So, I was saying they [the marks] are generated afresh [in each generation]. If you say ‘that is wrong’, to me you are saying they are not. If that’s an unfair characterisation of your position, I’d have expected the debate to have been far shorter.
The rest can be followed by anyone with the patience: post after post of you apparently confusing ‘evolutionary conservation’ with copying of the actual methylated state, and somatic replication of a methylated state with conservation across a meiosis/syngamy boundary. And me trying to explain. If my explanations were unrelated to your actual argument, why not simply agree?
This is ‘saying the opposite’?
What you said subsequently can be viewed by anyone with a scroll wheel. We may have misunderstood each other, but I did take great pains to explain what I meant, and you continued to disagree. Now you say you didn’t disagree in the first place.
Splutter! I suppose that’s as close as I’ll get to an acknowledgement. If you never made the argument, I wonder how so much bandwidth got consumed by you not-defending-it. A simple ‘agreed’ would have sufficed.
TomMueller,
Aha! I made it up! Rather, it is hypothetical. The known behaviour of remethylation is that it moves out from an already methylated region. So I would see it as analogous to an RNA primer. With a different ‘primer’ in males and females that survives the zygote wipe but not the gametogenetic one, being reset in the latter, we have a plausible mechanism.
The difficulty of supporting this hypothesis might be clear when one thinks of how imprints are found – by differential behaviours of particular genes when gender expression is interfered with. But such a ‘residual’ mark would be cryptic in genetic studies. Still, the difficulty of support works against the skeptic too – it is merely assumed that methylation is completely removed, because that is what is deemed to happen. It certainly is removed from genes, but how would you detect retained methylation in a small region of intergenic space?
I’m very skeptical about this. I don’t see how one can ensure that of the eight chromosome copies presented for potential representation in a grandchild, three fourths of which are thrown away (ditto any cytoplasmic factor), we can end up with a gender effect that survives through a randomly-gendered F1 (and up to two reciprocal crossovers).
Plus, it is being declared adaptive. Yet any potential adaptive benefit would be overwhelmed by stochastic factors. There’s the ‘75%’ problem, plus starvation generations would have to happen with significant, but periodic, frequency, to drive this. I doubt it would happen even then. I’d like to see a pop-genetic treatment.
I think the main determinant of evolutionary conservation will be the ‘stand-off’ I alluded to earlier. The full set of imprinted genes has a marked bias in favour of equal and opposing effects on placental growth. This constrains the evolution of either one of an imprinted pair. An imprint favouring males arises and fixes, but it selects for an imprint favouring females (a close analog of Fisher’s sex ratio theory). So that arises and fixes. Now they are stuck like that, bickering over resources for evolutionary time. One can’t move without the other, so their evolution is locked together and slowed, and hence conserved.
The naive expectation is that conserved = adaptive. Ain’t necessarily so.
Given the opposite effects seen in the Överkalix grandsons and granddaughters, declaring it “adaptive” does seem a mite premature.
[This is ‘meiosis’. It’s a figure of speech. Look it up, like I did.].
DNA_Jock,
Well, this could conceivably be an example of gene-level adaptation on different genes with opposing effects. That’s pretty typical of imprinted genes for example – see Igf2 and Igf2r in mice, which canel each other out, but are arguably ‘adaptive’ for the genders they individually favour. But it seems only barely possible to me, all considered.
Mung,
Damn, a smiley! You’re killing the ambiguity, man …
Not every thing inherited is by DNA inheritance! Some of the inheritance is epigenetic. I’ve said the mechanism of inherited imprints is not well understood. We know for some of the imprints the Xist lncRNA is responsible for X-chromosome imprints if the zygote is female so we can say that epigenetic mark is genetically driven, but some of the other imprints, not so well understood not the least of which reason we don’t believe we’ve found all the imprinted genes!
stcordova,
Like what? Imprinting failed you. Starvation studies are dubious. Scared mice?
I am talking of >1 generation effects. I am aware of no documented instance of a definitively characterised factor not derived from DNA, directly or indirectly, passing down multiple generations.
If an exception is found, I will change my tune – but simply to this: “The epigenetic marking goes via DNA inheritance, same as everything else except X.” . An exception does not become the new rule.
But first, find me an X.
It is well enough understood to be pretty sure that the genes performing the wipes and remethylations are conventional DNA genes. That’s how they come to exist – to be ‘conserved’ in organisms as diverse as mice, horses and humans. You can’t really suppose that we have an epigenetic relationship to the entirety of Mammalia, can you?
There’s a Nobel waiting for someone who can demonstrate such a bypass of Mendelian inheritance in mechanistic detail. Till then, your supposition is on equal footing to ‘it does not happen’, my preferred position.
Allan thank you for the conversation. You said at Noyau:
I phrase the idea differently, and so much so one may say I disagree. This is what I say:
I do agree the marks on the paternal DNA are de novo in the sense that they are added after the paternal DNA is erased as shown in block C of the photo of the OP.
But de novo marks using de novo methyl transferases 3(DNMT3) or whatever, does not imply the marks are not inherited, it only implies the mechanism of transgenerational inheritance isn’t achieved by using marks already present on the DNA as a template for some copy-and-paste-style method of marking such as the case in somatic cell epigenetic inheritance that is achieved by DNMT1.
The mechanisms of transgenerational epigenetic inheritance is not the same as somatic cell epigenetic inheritance.