Talk:Epigenetics/Archive 1

The article is a bit short on methodology, the methods and technologies used for determining the epigenetic state of a cell, the epigenetic equivalent of DNA sequencing methods. There is some in the intro but ideally there should be a dedicated section like in the "Genetics" article.Karlengblom (talk) 23:07, 2 January 2008 (UTC)

This is an example of name clash. There are two schools - one is a "nurture" school, which argues that the configuration of the carbon atom, the state of the cellular environment, the protein expression, and the womb environment, then through the adult, is all epigenetics. The only people who believe in the definition you cite here, are those who believe that genes contain all the information to express the phenotype, which is demonstrably not true (have you seen the pictures of the five different sheep cloned with Dolly?).

So, the environment-centric definition must at least be mentioned. Frankly, I didn't think the other one still existed, as I haven't heard it used by a serious biologist in that sense in a while. As I understood it, the old use of the term you cite here is now part of proteomics or genetics proper. Note the criteria for what is considered "epigenetic" in biology:

http://www.cshl.org/AnnualReport1999/rh5.html "Not all clones are created equal. As genetically identical cells (such as those in an embryo) multiply, different sets of genes are switched on and others off, giving rise to cells and tissues with distinctive properties (e.g., liver versus muscle).

Such differential gene expression is often established by alterations in the large-scale architecture, or chromatin structure, of DNA. ... Such states of chromatin are said to be epigenetic because they can be inherited in a stable manner."

This implies pretty damn strongly that anything that "can be inherited in a stable manner" is epigenetic. Your definition is the wrong way around - it tries to limit epigenetics to the expression, whereas the expression is only part of epigenetics, and considered so because of the "stable manner", not because of its relation to the physical structure of the genome or program... if gene expression depends on hormones or environmental stimuli, then those are epigenetic factors.

The open question leading to the debate implied is whether expression of the trait in the phenotype is what is important (as it is for sexual selection and ecological selection both), or whether it is the capacity for its being passed on to offspring as a latent (unexpressed or only partially expressed) gene that the field should study. That's a big difference between molecular and evolutionary biology.

Also, look at the use of the term "epigenetic" in every field outside biology:

There are dozens of conferences on epigenetic robotics, and unless you are going to argue that they believe that robots have animal-like genes, well... http://www.lucs.lu.se/epigenetic-robotics/ http://www.google.com/search?q=epigenetic+robotics

If you want to say that "epigenetics" is "the study of traits that can be inherited in a stable manner", and includes all forms of selection, natural and otherwise, and that "epigenetic expression" studies the biological influences, e.g. the chromatin structure, and that "epigenetic robotics" attempts to duplicate the behavioral and perceptual traits using non-genetic entities, I wouldn't argue with that. But proteomics is epigenetic too, by any definition I've ever heard, because the protein folding characteristics in cellular media aren't determined by genes but are stable across generations. So you're taking a position on the genetic expression versus environmental mediation by using the term in this very narrow sense.

My definition isn't "real" only if you believe in genetic determinism of the phenotype - which corporate geneticists believe in, and nobody serious does. 24

--- Various non-bullshit references showing the intersection of the ways the term is used. I don't see how there can be an "epigenetic view" of robotic outputs of biomimicry unless the word has a broader sense than molecular:

References: Sipper, M. et al (1997) "A Phylogenetic, Ontogenetic, and Epigenetic View of Bio-Inspired Hardware Systems", IEEE Transactions on Evolutionary Computation, 1(1).

Sanchez, E. and Tomassini, M. (eds) (1996) "Towards Evolvable Hardware", Springer-Verlag, Lecture Notes in Computer Science, 1062. Higuchi, T. et al (1997) "Proceedings of First International Conference on Evolvable Sys-tems: From Biology to Hardware (ICES96)", Springer-Verlag, Lecture Notes in Computer Science.

Beyond that, there are lists of biological references these people refer to: "A good exposition of current population Biology modeling is J. Maynard Smith's text Evolutionary Genetics. Richard Dawkin's Selfish Gene and Extended Phenotype are unparalleled (sic!) prose expositions of evolutionary processes. Rob Collins' papers are excellent parallel GA models of evolutionary processes (available in [ICGA91] and by FTP from ftp.cognet.ucla.edu:/pub/alife/papers/ ). As fundamental motivation, consider Fisher's comment: "No practical biologist inter-ested in (e.g.) sexual REPRODUCTION would be led to work out the detailed consequences experienced by organisms having three or more sexes; yet what else should[s/]he do if [s/]he wishes to understand why the sexes are, in fact, always two?"

A full and proper biological epigenetics would probably have to explain why two sexes evolved, but three never did in any species. That's not necessarily going to be coming from the chromatin, although it might...

The word "epigenetic" has nothing to do with "epigenetics". It comes from "epigenesis". There is no name clash over "epigenetics". AxelBoldt

• There were some biologists floating around giving lectures to the contrary in 2000. It was from one of these that I learned the definition I used... maybe he was *arguing for* this broader definition and it was original to him? Hard to tell standard from novel use when the work is state of the art - as I recall this was a single biologist explaining the phenomena so that the robotics folks could follow to explain how their biomimicry worked... maybe the moleculars won a terminology fight some time back, or maybe this more behavioral biologist was being way loose with epigenetic vs. epigenetics... whatever. It appears the narrow use is universal and the other is not. Perhaps if you related proteomics to this narrower and strictly-molecular definition of epigenetics, it would be clearer... perhaps a new name has been placed on the field studying the whole environment... However, I suspect strongly that this is reductionism by the molecular boys, trying to find clues to everything in little bits of DNA, rather than looking at stuff like wombs and environment conditions... 24

Should we mention Trofim Lysenko in this context, or is that too far-fetched? --Magnus Manske, Monday, 13:48, 15 April 2002 (UTC)

http://www.pitt.edu/~sshostak/

Waddington should be credited with injecting 'et' into epigenesis, turning it into epigenetics and drawing attention to the role of genes in development.

Harvey might also be mentioned by way of coinage, but he credits Aristotle, the Philosopher, with originating the concept of organic unfolding. See Shostak, S., "Death of Life: The Legacy of Molecular Biology." London: Macmillan; 1998.

— Preceding unsigned comment added by 67.165.107.201 (talk) 12:06, 16 January 2004 (UTC)

Some scientists at Cancer Research UK and Cambridge's Babraham Institute think enzymes stablize epigenetics

That would be canalisation? - Samsara 16:42, 1 January 2006 (UTC)

The articles to be merged are epigenetics and epigenetic inheritance. There is a difference between these subjects in that the former is a field of study while the latter is the phenomenon itself. Therefore for clarity I think that both pages should remain separate. There will be some information more appropraite for each page, and there may be some overlap, but Wikipedia is not paper. Bensaccount 17:40, 18 February 2006 (UTC)

On second thought, I think that a merge might make organizing this easier. Bensaccount 17:55, 18 February 2006 (UTC)

There are compelling evidences showing that events such as holocaust and 9/11 have impact on the next generations

see the discusstion at

http://jcem.endojournals.org/cgi/content/full/90/7/4115

http://www.peakoil.com/fortopic18784.html

http://www.bbc.co.uk/sn/tvradio/programmes/horizon/ghostgenes.shtml

I wish to see these references are to be cited here

— Preceding unsigned comment added by 211.31.126.226 (talk) 11:52, 9 July 2006 (UTC)

I second this. I've come here from seeing the BBC Horizon programme. I hoped to learn more, more history, current state of the research. Instead I found myself a little lost.

— Preceding unsigned comment added by 70.71.39.12 (talk) 05:54, 20 January 2007 (UTC)

I apologize if I stepped on any toes with my last revisions - I've deleted a considerable amount of text that was in my view redundant within the article or which should be referred to other Wikipedia articles referenced in the text, while making some additions. Some points on which you may disagree: I removed text suggesting that mitochondrial DNA mutations are epigenetic, because I don't think that's a typical usage; I distinguish epigenesis from epigenetics as an early 18th-century theory opposed to preformationism; I accept the idea of epigenetics as "anything but DNA" even though this practical distinction is historically fairly modern. As the Wikipedia help says, "be bold" ... sorry if you need to revert something! Mike Serfas 02:22, 18 August 2006 (UTC)

The is the stub article Bookmarking related to this? Bookmarking has very few links, after I fixed the ones looking for bookmark (computers). Someone who knows the difference between epigenetics and epigenetics should (heh) should throw a few links at it. -- Kendrick7 22:12, 10 October 2006 (UTC)

Although my degree in Biology is over 20 years old, I should be able to understand this article. I found it to be very hard to read. —The preceding unsigned comment was added by Oswalia (talk • contribs) 16:47, 11 December 2006 (UTC).

• I concur. I have tagged this article with a request for consistency in reference formatting as I feel this is its worst failing, but the style could do with improvement in addition. StoptheDatabaseState 14:40, 13 December 2006 (UTC)

I think both of your comments are too kind. The opening section, to say nothing of the rest of the article, is completely incomprehensible for me. Someone like me, who has taken a college-level biology survey course, should be able to at least understand the opening of an article like this, if not the entire article (Wikipedia is not a biologists' Wiki). Since the topic was discussed recently in Discover magazine, it must be possible for reasonably intelligent lay people to understand this topic on some level. As it is, the article is nigh worthless since it means nothing to people who don't already understand the subject. It needs heavy revision, IMO. 205.157.110.11 04:49, 25 January 2007 (UTC)

• I agree with the above comments. It is both welcome and essential that experts in the field contribute their knowledge in what is evidently a complex topic, currently apparently in the research phase, but the opening paragraph, at least, needs to give the layman some understanding. By the time I got to the 5th word in the opening paragraph I was lost. The wikilinks help, but they go to pages with the same issue. Could the meaning be stated in a second way? Could examples be used? Scanning further down the article I saw reference to a half-liver half-intestine cell, which suggested to me that this must be about understanding how some cells develop livers and some develop into other organs etc.--Rye1967 10:07, 20 February 2007 (UTC)

• This article is not suitable for Wikipedia due to its unreadability and total lack of attempt to reach the audience (almost all of whom will not be specialists in this field, but generalists who wish to learn more). It needs a complete rewrite by another author. [Frankly, even though many technical articles are written with a similar degree of incomprehensibility, there is no need for it, as Einstein and Galbraith have famously remarked]. I have a PhD in science from MIT, so am not unversed in technical reading in a wide range of disciplines, but this is over the top. Hambleton (talk) 18:37, 25 July 2008 (UTC)

I have to confess that I'm troubled that the author has not taken the above concerns more seriously. I'm in the neurosciences, and I hear all the time how difficult it is for people to come across material that only experts can read. Scientific writing has to start with the principle that there is a necessary hierarchy of abstractions much like a pyramid, and if the base has not been established for the reader, all the fancy stuff on top is simply going to appear as 'gibberish'. There are so many terms here that are undefined, so many concepts that are unexplained, and the author assumes throughout that he is talking to a peer, a fellow expert, when far and away most people dialing into this page are simply looking to learn something of the basics about Epigenetics. Unfortunately for that vast 'unwashed' majority, this piece is virtually indecipherable. As a PhD with a modest level of sophistication about these issues (but no real expertize) I found this piece frustrating. This indeed is one of the basic problems facing the entire Wikipedia community - how to present science in a state of the art fashion and at the same time make it accessible. It is really worth reminding ourselves what Einstein said about scientific writing, namely, that if you cannot explain your scientific ideas to a nine-year-old, you need to go back to the drawing board. This suggests that this author, despite some impressive mastery of the literature, needs to go back to the drawing board. DFW Harvard Medical School —Preceding unsigned comment added by 74.75.22.221 (talk) 01:13, 24 June 2008 (UTC)

This article has been quoted and reviewed on 'Epidemix' and later featured on Wired. In the article, it is suggested that the first paragraph (an probably several other areas of the article) are too advanced for the average reader. --TheTallOne 07:53, 12 May 2007 (UTC)

I work in the epigenetics field, have an advanced degree and the first paragraph does a poor job of defining what epigenetics is. Epigenetics is defined as stable, heritable changes in gene expression that don't involve changes in the primary DNA sequence (ATCG) itself. These modifications include methylation of cytosine base and methylation, de/acetylation and/or ubiquitination of histone proteins which associate with the DNA. DNA and histones make up what is called chromatin. These modifications to the DNA or to the histones influence chromatin structure in such a way to either allow or prevent accessibility of the proteins responsible for gene expression. These changes are heritable across cell divisions (ie daughter cells have the same epigenetic pattern as the parental cells) and reproduction (epigenetic patterns are reestablished during embryogenesis based on the original epigenetic pattern of the gametes). I think the definition of epigenetics can be best understood contrasted with the more familiar "genetic" changes, such as mutation of the DNA. Most people understand what a mutation is, whether it be a point mutation, deletion, or some kind of translocation. Genetic alterations change the primary DNA sequence and unless they occur in gametes, won't be inherited by the offspring. Epigenetics involves the DNA/Chromotin. I've never heard any definition which involves protein modification by anyone in the field. I think its important not to out think ourselves and define the concept as it is generally used in the field. --Mawst95 20:42, 12 May 2007 (UTC)

Why not replace the Introduction in the article with what Mawst95 wrote above? Starting with "Epigenetics is defined as..." Psoreilly 21:50, 13 May 2007 (UTC)

I think it is misrepresenting Winston to link to genetic determinism when the lecture really was about determinism. Winston mentioned Dawkins book/idea, The Selfish Gene but failed to advance Dawkins' views on the myth of genetic determinism which are better spelt out in The Extended Phenotype but we must assume he was aware of Dawkins' position on genetic determinism. I guess that subtle omission pales when Winston advanced a timeline of Darwinism to Eugenics and forced sterilization and then the Auschwitz concentration camp thus trying to paint some scientists and writers in colours both unethical and anti-semitic. Winston could do much to advance the public understanding of Epigenetics (which to most of joe-public the closest they have heard of is with prions in Mad Cow Disease) and this lecture isn't it. But then the public wasn't the target audience anyway. Ttiotsw 09:24, 14 May 2007 (UTC)

The whole Philosophical implications section should be removed. It seems to me to have been added solely as a stick to beat "determinism" (whatever that is) with. It adds nothing scientific or philosophical to the article, and is sourced primarily to two Winston pieces, one a lecture, the other an interview. The nod to Bateson is precisely that, a nod; while the quote of Jablonka's work is out of context, it is from a section not strictly about epigenetic mechanisms (more about (mis)perceptions of genes). Returning to Winston's remarks (or, at least, the reported versions here), they are very unclear, at times worryingly so ("some inherited genetic variations are not random"???). Furthermore, blandly stating that "calls this an argument against the 'selfish gene'" overplays the strength of the sources here. --Plumbago 16:18, 15 May 2007 (UTC)

Yup. Whatever that "philosophical implications" section is there for, it certainly does not address any philosophical implications... Snalwibma 16:22, 15 May 2007 (UTC)

Perhaps our mutual friend NBeale could answer this. I just love it when people base arguments on "determinism" as it it trivial to destroy. It is like grist between the two millstone of 3-body problem and emergent behaviour (just plucking two stones from my rockery). Ttiotsw 18:02, 15 May 2007 (UTC)

Looks like he has bigger fish to fry for now. Anyway, let's look at some of the statements in the offending section.

• genes are not the only unit of heredity

And? Is this statement disputed by anyone? I accept that it might come as a surprise to non-scientists though.

• some inherited genetic variations are not random

What exactly is meant here? Is mutation not random? Does this statement mean: a) that not all genes mutate at the same rate, or b) mutation is directional (a "guiding hand" perhaps)? This statement is too important for its ambiguity to be left as is.

• environment has a powerful effect on how genes work

High school biology triumphs again. The Extended Phenotype provides a nice intro to this.

• acquired information can pass onto the next generation i.e to say Lamarckism is not the necessarily wrong

True, but only in a limited way. Proper Lamarckian evolution is more or less precluded (c.f. The Blind Watchmaker for a nice intro), but there are well-known cases where Lamarckian-like processes occur (e.g. prions).

• evolutionary change is not necessary due always to natural selection

Again, high school biology triumphs. When I was at school I even wrote a computer program to simulate genetic drift.

• It is accepted by working geneticists that "the popular conception of a gene as a simple causal agent is not valid" and the gene cannot be seen as an autonomous unit

This is accepted by just about everyone! (Dawkins included NBeale) Everyone knows that there is a yawning gap between the genetic code and a specific phenotypic effect; a gap that's filled/obscured by all sorts of environmental factors (many of which are, of course, the product of other genes). It's certainly the case that the width of this gap is variable and not well-characterised in most cases (although in others it can be fairly straightforward, e.g. GM luciferase), but it's clearly also the case that despite "environmental effects" organisms tend to develop in the same way time and time again.

Anyway, beyond confusing things, I really can't see the purpose of this section (or, rather, I think I can ...). I'd remove it now, but I'd prefer some feedback first before doing so. --Plumbago 08:34, 16 May 2007 (UTC)

• Yes - as you say, it is a section devoted to knocking down a selection of elementary misconceptions and straw men, really just a list of obvious points which does not by any stretch of the imagination live up to the "philosphical implications" billing. The whole of what is in this section could in fact be reduced to a simple statement added to the very first sentence on the article - something like "Inheritance does not depend solely on genes". But in effect it already says that (though perhaps not quite as starkly as that!). So I reckon it can all be deleted. But I too will refain for a while from doing this. I'd like to hear what others think. Snalwibma 10:03, 16 May 2007 (UTC)

OK. No further discussion. I'll remove it. --Plumbago 21:26, 18 May 2007 (UTC)

Sorry guys - I've been v busy and only just had a chance to come to this. It could be argued that any attempts we might make to draw philosophical implications were OR, but the fact is that Robert Winston certainly suggests in his lecture that the fact of epigenetic inheritance has significant philosophical implications - he singles out Dennett and Pinker as egregious examples of people who write about genetics and seek to draw philosophical conclusions without knowing what they are talking about. And his main point is that the idea that our understanding of science and genetics is nearly complete (bizzarely pushed by Dawkins in his Times interview) is simply wrong. Furthermore the "Evolution in 4 Dimensions" book also specifically draws philosophical implications, and I think we should reflect these. It got a v good review in either Nature or Science. NBeale 21:31, 18 May 2007 (UTC)

Twas Nature 435, 565-566 (2 June 2005) by Massimo Pigliucci no less! NBeale 21:39, 18 May 2007 (UTC)

BTW the summary of Winston that was in the article was not at all the one I wrote, but someone elses which I thought rather destroyed the force of what he was saying. NBeale 22:02, 18 May 2007 (UTC)

OK. Can we drop Winston already? Epigenetics is not his primary field, and though he may discuss it in his lecture, his sources are far from the best for a technical article on it. Anyway, the material you've re-added about Winston is basically to do with Dawkins (yet again), so at best is only appropriate over there. Regarding epigenetics' "philosophical implications", you need to spell them out properly here rather than merely allude to them. And why aren't they "technical implications" anyway? To my mind, epigenetic inheritance is simply an additional mode of inheritance, with no obvious "special" significance beyond it contravening earlier models of inheritance. It has really rather limited effects (as the article already states). And as for statements like "Thus the simplistic view of evolution solely in terms of 'selfish genes' ...", well, they are referring to a straw-man and have no place here (or elsewhere). --Plumbago 06:45, 19 May 2007 (UTC)

The mechanism is simple. Paternal effect because the nuclei of eggs are formed before birth and never change afterward. Thus eggs do not reflect later epigenetic changes during the mother's lifetime; whereas sperm cells form constantly and reflect the current state of modified epigenetic DNA. Delayed until grandchildren because both halves of the chromosome must have the epigenetic modification in order for the base gene behavior to be completely overridden on both halves of the chromosome. Since only the father passes on new epigenetic information, his children will not reflect any such changes requiring both halves of the chromosome to be modified by epigenetics. Only grandchildren who get the same epigenetic change information from both grandfathers will have the epigenetic change. Of course maternal effects do exist from that narrow time before the female fetus forms all her eggs. Very very elementary Mendelian stuff.. 69.23.124.142 (talk) 19:33, 16 September 2008 (UTC)

I have removed this, which was in ref tags after "maternal effects":

However Robert Winston cites a study Vaigron in North Sweden of families in the late 1800s and early 1900s which showed that the paternal (bur not maternal) grandsons of people who had been involved in famines were four times more likely to develop diabetes. Lecture at University of Dundee)

I haven't seen the original study, but in the absence of anything in this article to explain otherwise, I can see no reason why this must be caused by "paternal effect" rather than imprinting, therefore I can not verify that it is relevant. Joe D (t) 14:33, 15 May 2007 (UTC)

The ref is "But Paternal effects are not unknown: Robert Winston mentions a study by Pembry & Bygren (European Journal of Human Genetics 14 159-166) of families in North Sweden the late 1800s and early 1900s which showed that the paternal (but not maternal) grandsons of people who had been involved in famines were four times more likely to develop diabetes. Lecture at University of Dundee)" I am relying on Winston's summary, I haven't seen the original article either, but the fact that it was paternal and not maternal grandfathers that mattered strongly suggests a paternal effect. PS: looking at the discussion at Leeds University I think it's quite clear that this is a paternal effect, so I'm putting it back. NBeale 06:17, 16 May 2007 (UTC)

Did he actually say "paternal effect"? It could very easily be imprinting, and that seems a far more likely explanation to me. Joe D (t) 06:37, 16 May 2007 (UTC)

I have amended the reference so that it includes a link to the Pembrey article: PMID 16391557 - and we can go and judge for ourselves whether it deserves mention as a possible paternal effect without relying on Winston's interpretation. What I don't understand is why Winston needs to be mentioned at all. I have stopped short of cutting him out, but it would seem reasonable to do so, even if the ref stays. Snalwibma 10:39, 16 May 2007 (UTC)

I see no objection to removing the Winston ref now that we have the primary one. Originally I have got the name phonetically and Bygren became Vaigron!! NBeale 22:40, 16 May 2007 (UTC)

Excellent! I was indeed totally mystified by "Vaigron"! Snalwibma 06:12, 17 May 2007 (UTC)

Does anybody have the full text of the original article? My library isn't subscribed to the journal so I can't get it. I read the "comment" article that went with it though, and it was pretty clear that the mechanism was, at the time, unknown, but states that it's likely just to be another case of imprinting -- a slightly interesting twist to an established mechanism. I've yet to see a reference that states that this isn't the case. Joe D (t) 23:01, 16 May 2007 (UTC)

• Full text? Try this. It seems to be available without subscription. If you have problems, send me a message on my talk page and I can email you a copy. Snalwibma 08:44, 17 May 2007 (UTC)

Thanks, I guess it wasn't a subscription issue at all, just a problem with Nature's DOI redirecting thing. Anyway, this bit stood out:

Little is known about the transmitting mechanisms, except in the specific study of epigenetic states at the AxinFu allele, and the altered DNA methylation patterns at two loci in the epididymal sperm of rats with reduced spermatogenic capacity due to their paternal ancestor being exposed (in utero) to the endocrine disruptor vinclozolin.

So, it appears to be a kind of ad hoc version of imprinting, not an effect of paternal gene expression. Joe D (t) 22:28, 18 May 2007 (UTC)

Actually, the more I think about it, the more obvious the irrelevancy of this study seems. I wonder, does NBeale actually know what "maternal effect" is? Joe D (t) 22:34, 18 May 2007 (UTC)

How about something like this. "It used to be thought by working scientists, and is still popularly supposed, that biological inheritance is nothing but genes (possibly even "selfish" genes) and DNA. It is now understood that, in addition to the fundamental roles of genes and DNA, there are other mechanisms of biological inheritance. In Biology, epigenetics is the..." NBeale 21:59, 18 May 2007 (UTC)

Hmmmm. It certainly shouldn't go at the very start. This is an article about a biological subject; heading up with some statement about the historical position of epigenetics overplays things. Especially an overblown statement that appeals to popular misconceptions of a technical subject. However, I would certainly agree that the article intro is a bit technical (and long). It might be an idea to write a new semi-technical intro, and consign the existing intro to a proper "Introduction" section. Among other things, it would (to my mind) help to just point out that epigenetic inheritance deals with several distinct mechanisms by which information is passed inter-generationally outside of the genome. The part about altering genetic sequences in the opening sentence, while correct in itself, is misleading it seems to me (and omits consideration of entities like prions). Maybe, "In biology, epigenetics refers to a class of mechanisms by which heritable information outside of the genome is transferred from parent to offspring organisms." Too clunky perhaps? --Plumbago 07:09, 19 May 2007 (UTC)

I propose something along the lines of:

In Biology, epigenetics is the study of all heritable and potentially reversible changes in genome function that do not alter the nucleotide sequence within the DNA. At the molecular level, the majority of biological processes are carried out by proteins, collectively the proteome, while the structure, and thus function, of these proteins is determined by genes, collectively the genome. The proteome changes in response to the cell's environment, thus allowing it to perform various functions, and this is most obvious in multicellular organisms, where different cells perform different functions. The genome, however, is static, and epigenitic mechanisms are thus an important mechanism of "cell memory", allowing cells to differentiate and cooperate in multicellular organisms. Additionally, some epigenetic inheritance is trans-generational, and thus may be subject to evolution, including the potential for the genome to be modified in response to the environment in germ-line cells.

Because:

1. what the article is about should be established in the first sentence,
2. epigenetic inheritance is very important in development and differentiation, but only an interesting side-note in evolution.

However, I can see a potential issue with it: we're talking about the implications before really mentioning the mechanisms. The implications are indeed important though, and I imagine most laypeople (and indeed, developmental and evolutionary biologists) will be more interested in the implications than the mechanisms.

Joe D (t) 16:27, 19 May 2007 (UTC)

One problem I have with the current intro is that it rather narrows the definition of epigenetic inheritance to genome interactions. These are, to be fair, the predominant epigenetic mechanisms, but there are other processes that are examples of epigenetic inheritance that do not directly involve the genome at all. For instance, some proteins self-template such that functional forms can only be assembled in the presence of existing functional forms which are inherited from parent organisms (e.g. Cheng, M. Y., Hartl, F.-U. and Horwich, A. L., 1990, The mitochondrial chaperonin hsp60 is required for its own assembly, Nature 348, 455-458). This sort of structural inheritance is implied in the case of infectious prion agents, and more generally applies to cytoplasmic properties (e.g. Meyer, E. and Beisson, J., 2005, Epigenetics: Paramecium as a model system, Médecine/Science 21, 377-383). My point in raising this is that perhaps the first sentence should be altered to something like, "In Biology, epigenetics is the study of inheritance that occurs via mechanisms separate from traditional genetics", and then follow on with some examples. Otherwise it seems like we're missing a trick. Anyway, as this is pretty far from my own field (to say the least!), I'm likely to be out of date here. Cheers, --Plumbago 09:12, 21 May 2007 (UTC)

P.S. Centrioles have been thought of as self-templating structures that represent an example of epigenetic inheritance. However, a paper published just this past week in Science finds instead that they can be synthesised from scratch (using genetic information), suggesting that they don't represent inheritable epigenetic information after all. The full citation is: Rodrigues-Martins, A., Riparbelli, M., Callaini, G., Glover, D. M. and Bettencourt-Dias, M. (2007) Revisiting the Role of the Mother Centriole in Centriole Biogenesis, Science 316, 1046-1050 (DOI: 10.1126/science.1142950). --Plumbago 12:22, 21 May 2007 (UTC)

Hi Plumbago - thanks for the ref. But I don't think this is a correct inference. The paper (which doesn't mention epigenetics at all) says that "The mother centriole is not a bona fide template but a platform for a set of regulatory molecules that catalyzes and regulates daughter centriole assembly." but this is by no means incompatible with the idea that alterations in the mother cenrtiole will be inherited by daughter centrioles. No-one suggests that epigentic effects would operate in the absence of genes, merely that they are another, important, dimension and the popular misconception that DNA=Genes=Biological inheritance is simply wrong. NBeale 20:55, 21 May 2007 (UTC)

Regarding the centriole paper, it does specifically address the issue of self-templating. Possibly not in the abstract - I had a look at the paper too. And from my old biologist days, I'm pretty sure that centrioles were cited as an example of "structural" (= epigenetic) inheritance. They might still be - the paper is only looking at them in a model system (Drosophila). There are other structural examples too - the pattern of cilia on the surface of Paramecium as I recall.

On the other point, I think we all agree that "genes = all biological inheritance" is incorrect, but that's not to say that "genes = 99% of biological inheritance" isn't correct. As it stands, epigenetic mechanisms don't appear to provide anything near the "bandwidth" of genes. Also, the epigenetic mechanisms we do know about appear a very crude means of transferring information between generations. And on top of that, as I read things, some epigenetic mechanisms may ultimately be genetic in basis, but transferred between generations by epigenetic means (i.e. imprinting; the extended phenotype). For all these reasons, I'd argue that it's overplaying things to make the sort of strong statements in the "philosophy" section - the "genes = biological inheritance" line is a straw man, plain and simple. --Plumbago 21:41, 21 May 2007 (UTC)

The evolution subsection of the functions/consequences section, previously entitled "Epigenetic effects and evolution" somewhat misses the major point in my opinion. The section is mostly concerned with somatic epigenetic inheritance, leading to the differentiation of cells. It goes into detail of trying to justify this within the existing model of the modern synthesis, for two paragraphs. I can justify it in one sentence: somatic epigenetic inheritance is no different from the rest of the proteome and mechanisms of expression regulation in that it is a product of the genotype. Chromatin remodelling, for example, is carried out by HATs, HDACs and remodelling complexes with the aid of transcription factors, morphogens, cell signalling, etc, and needs no more justification or special treatment than glycogen production or the TCA cycle. Where evolution is interesting is in the germ-line, and transgeneration epigenetic inheritance. Does anybody have any objection to an overhaul of that section? Bits of it may be worth incorporating into the development section? Joe D (t) 17:19, 19 May 2007 (UTC)

If you are interested, you might want to look at the reelin page. I'm not a specialist in genetics, let alone epigenetics, but thought that this may interest you. Best regards, CopperKettle 18:33, 19 May 2007 (UTC)