From my understanding of epigenetics, even though there is inconsistency in inheritance of there marks, there is another possibility that they might skip a few generations and reappear again. In addition, cells also initiate repair mechanisms after a certain time and the question is how and due to which factors.

Epigenetic modifications on the histone level are assumed to be mitotically stable, for many generations. Moreover, recent studies have indicated that epigenetic modification can be even stably transmitted through the germline to the next generation of human beings.

This may be different when looking at cells in culture, because they should adapt according to treatment (medium, FCS, antibiotics, etc.), but it would make me wondering, if epigenetic modification are stable only for 1 cell division.

Personally, I assume histone modification changes as fast adpator mechanism to a changing environment, while genetic according to Mendel will take more time.

There is some evidence also from the mouse system, that changes are occuring on the epigenetic layer within a few generations. This is e.g. the case with obese mice. Another good example for transgenerational epigenetics is liver enzyme expression. There was a recent study feeding male rats with a certain kind of food that caused the expression of additional liver enzymes. Such male mice were bred with normally-fed female rats, and the offspring expressed again the strange enzyme pattern of the father. Thus, epigenetics may change upon purpose (adapation), but should normally keep the status quo.

For an updated general overview on the issue, this paper should be helpful, especially with the references:

http://www.cell.com/trends/ecology-evolution/abstract/S0169-5347%2812%2900049-3

http://www.bonduriansky.net/TREE-2012.pdf [Author's personal copy]

"...Recent discoveries in molecular and cell biology have revealed novel mechanisms, such as transgenerational epigenetic inheritance, that could account for a variety of parental effects..."

There's actually some new work that came out recently describing the mechanism of this inheritance, and wasn't what I expected: http://www.sciencedirect.com/science/article/pii/S009286741200935X

Apparently, it's not the histone modifications themselves that are inherited, but the proteins that put these marks on. Granted, this paper only describes two marks and two modifiers, but I think it's a big step toward understanding the degree of heritability of these things.

@Brian: Right, and that is consistent with the mitotical stability: the "informations" of "where" and "how" its logically not in the markers themselves but in the proteins cascades that place them.

personally, I believe this topic is really fascinating! There are evidences that, in microorganisms, heritable epigenetic alterations underlie most phenotypic changes, and it is likely that a similar heritable effect of environmental exposure occurs also in mammals. In addition to the articles cited above I've found some articles suggesting that microRNA molecules transmitted through meiosis can restore the epigenetic state in zygotes.

http://www.sciencedirect.com/science/article/pii/S1534580708001196

http://www.nature.com/nature/journal/v441/n7092/full/nature04674.html

@Linda: Indeed it is really fascinating! Have you heard about "the first evidence" of "reversible epigenetic changes associated with behavior" in bees?

You read well, I wrote "behavior" and "reversible"!

http://www.nature.com/neuro/journal/v15/n10/full/nn.3218.html

The study it's mainly about DNA methylation, so it's a little bit out of Kristopher question, but it can help to focus on the scope of the field..

I remember reading a paper a few years ago mentioning oocytes passed female epigenetic pattern to the human embryos. And according to another paper, even our food consumptions effect our epigenetic pattern (even our behavior) because of blood-streaming plant miRNAs.

Summerizing all the answers, "Most of the epigenetic pattern inherited throughout generations within cells in the same environment"

My opinion;

- Epigenetic pattern created by certain enzymes such as DNMTs and HDACs; so during S phase, these already existing proteins marks the new created DNA same as the original. Which keeps the daughter cells the same way.

- Small changes of environmental inducers may change this pattern slightly, even in the same conditions like an overnight incubation. So next generation is not the perfect copy of the previous one.

- More dramatic changes occur during germ cell generation and embryogenesis which causes the erase and rearrangement of previous pattern. But embryo still carries some pattern of the parents, especially mother's.

You can track down changes of certain modifications throughout generations of lab animals or passages of cultured cells and that would be a good study!

Histone deacetylase (HDAC) inhibitors increase histone acetylation and enhance both memory an synaptic plasticity. The enhancement of hippocampus-dependent memory and hippocampal synaptic plasticity by HDAC inhibitors is mediated by the transcription factor cAMPC resonse element-binding protein (CREB). The HDAC inhibitors enhance memory processes relating long term behaviorals by the activation of genes reglated by the CREB:CBP transcriptional complex , them the histone modifications is realy a true epigenetic factor that inheritance for several generations.

Honestly, one of my prospective advisory committee members does not believe histone modifications are epigenetic. In fact, it is suggested that microRNAs are more epigenetic than HM's. In my own opinion, I agree with this assessment, but must add that DNA methylation is probably inherited.

It does not matter how many generations these modifications can be inherited, they are not part of the GENETIC code which is solely encoded by your DNA and is 100% stable, besides spontaneous mutations which are usually corrected, thus anything else is considered Epigenetic!

@Kevin et al., we noted multigenerational and transgenerational inheritance of global methylation patterns and Dnmt expression, associated with increased mammary cancer risk:

http://www.nature.com/ncomms/journal/v3/n9/full/ncomms2058.html

Position effect variegation (PEV) is an excellent example of this type of epigenetics at work. In the lab in which I did my PhD (now quite awhile ago....), it was found that placing a gene next to a telomere could case reversible gene silencing (Telomere Position Effect - TPE). This work was done in the model organism S. cerevisiae, which does not have microRNAs or DNA methylation. Histone modifications are the mechanism for silencing. The link shows a cartoon of a yeast colony with sectors (white yeast = gene on; red = off). The sectors show many generations of heritability, with the gene switching to either the on or off state. This also can be nicely visualized in the Drosophila eye.

http://dx.doi.org/10.1038/npg.els.0006040

Not all histone modifications are epigenetic and imprint some type of inherited cellular memory. The idea of heritable cellular memory comes mostly from genetic studies in flies, where Trithorax and Polycomb proteins were responsible for maintaining patterns of gene expression during development. Many of these proteins are histone methyltransferases or associated co-factors. Histone acetyltransferases were not identified as "epigenetic" mutations, at least in the unbiased genetic screens.

There is significant debate about the mechanisms of inheritance of methylated histones during DNA replication. Recent work From Alex Mazo showed that histones are replaced during DNA replication but the Polycomb proteins remain associated with the replicating chromatin, suggesting that new nucleosomes are methylated as they are assembled or shortly thereafter.

Is correct, not all histone modifications are epigenetic changes, but histone methylation too related memory formation, it is a actively regulated in the hippocampus and facilitated long term formation related long term behavioral, them it is another histone modifications's epigenetic factors.

Its a good question to ask and debate because there are many things and conditions that can disclose this fascinating but interesting topic. In relation to the question asked "Up to how many generations these modifications can be inherited", this all depends upon histone modifications that help modifying the genome like methylation status which decides whether the genes will express or not, there are many mechanisms of Histone modifications and each works differently. These methylation marks that develops during a life is heritable or not depends upon the levels. Earlier it was supposed that Methylation marks got erased and then with the help of certain denovo methylases recreate methylation patterns which are then maintained by maintenance methylases but now some research in the field are challenging the erase of methylation marks and says that they remains same and preserved by denovo methylases which is may be through preserving the histone modifications which is all the same but hypothesised differently as both mechanisms works simultaneously and affects each other. So these changes keeps on moving and are inherited to generations but each generation contributes a change in modification whether it affect a phenotype or not.

For what I've read on the subject I believe that at least one point can be set: there should be a difference between the terms that define epimutations resulting in a changing of the DNA nucleotide sequence and those which have other consequences in regulation, expression, etc. This would at least help to distinguish among two main types of epigenetical changes and then their general characterization. Something like "interactive epimutations" (i.e., histone tails chemical modifications) should be different from "altering epimutation" (methylated cytosines converted into thymines).

Interesting: lncRNA have a role in epigenetic changes direclty during replication!

http://www.the-scientist.com/?articles.view/articleNo/32637/title/Lamarck%20and%20the%20Missing%20Lnc

"...lncRNAs appear to bind to transcripts in the nucleus as they emerge from the replication fork of the DNA, and recruit enzyme complexes to induce epigenetic changes at these loci..."

About what is an epigenetic processes in the CNS: J. David Sweatt in "Experience-dependent Epigenetic Modifications in the CNS" (Biological Psychiatry 65(3):191-197, 2008, p. 2) suggest "...dynamic regulation of epigenetic mechanisms is a part of the normal gene-environment interface..." a simnple one how these and in the same time, a complex mechanism.

With all the buzz about epigenetic the current usage of the word (however incorrect) is applied pretty much to any transcriptional regulation since transcriptional regulation involves a change in what is referred to as "epigenetic signature" - another misused term. In terms of linguistic epigenetic is a horribly unspecific word since it refers to all genetic information other than the information contained in the DNA sequence. In my opinion we need to come up with a more stringent nomenclature for the many different mechanisms currently referred to as epigenetic modifications.

In regards to the longevity of epigenetics: as a rule of thumb direct transcriptional activation can change the "epigenetic signature"rather rapidly (within an hour). This mode of regulation is rapidly reversed. Epigenetic reprogramming of histone modifications (e.g. by inhibition of the modifying enzyme) occurs in approximately a week. In contrast, the more persistent DNA methylation can be reversed in approximately one month. That said, these are just loosely defined guidelines.

In terms of persistence (and the original definition) epigenetic modification should be maintained in the offspring. Working with cell lines this is hard to define, but is often considered as maintained through cell division (through a yet unidentified mechanism).

In vivo, the c-kit gene is an established model of epigenetics. It was shown that the inheritance of the phenotype relies on a RNA, thus putting RNA as a carrier of epigenetic information transversing generations (Nature. 2006 May 25;441(7092):469-74.).

How alltered histone modifications are considerate epigentic marks? Kinds and diverse effects:

Of course there are very kind of altered histones relating anything else that synaptic changes about memory function, but I will refer only that. However, altered histone is a very broad activity of dynamic regulation of gene expression and the gene-environment interface is apply to specific memory activity with appetitive & adversative circuitres to long term behavioral, and for example, the recovery memory funciton subserve for a window time a modifications to the autonomic response, affecting the reactivity neural circuitres, Though these one conerning a subjetive perception above the accumulate experiences, it has certificated the epigenetic mechanisms promote the storage of memory because they can stably contro gene expression over long periods. The environmental changes endocrine system's response links the memory activity (vertebrate & invertebrate) with the epigenetics marks. These one have an important role of synaptic plasticity and memory formation, to these depends what gene expressed. In this function, for example, altered histone methylation regulating memory formation, and histone inhibitors deacetylase enhance memory and synaptic plasticity, at the last, altered histone acetylation is associates with age-dependent memory impairment.

mistake typewrites:

where say: the recovery memory funciton, must say : the recovery memory function,

whehe say: Though these one conerning, must say: though these one concerning,

where say: can stably contro gene expression, must say: can stably control gene...

J. HSIEH & GAGE, FH propose in the conclusions in "Chromatin remodeling in neural development & plasticity" (Current Opinion in Cell biology 17:664-671, 2005, p. 670): "...We propose that chromatin remodeling conveys dynamic environment experiences to static genome, resulting in heritable & reversible changes in the CNS during development and throughout adulthood.."

Samarjeet Singh indicated that epigenetic marks acquired during a life time can be inherited. This is analogous to Lamarckism, which was out favored by Darvinism. I am wondering if there are solid data to support the notion that acquisition of changes during a life time can be passed on to the next generation. No matter whether this is due to DNA methylation or histone modifications.

To Sundararajan Jayaraman:

My answer to your request for solid data to support the notion that acqusition of changes during a life time can be passed on to the next generations. At first, any reference to concepts: Wolf, M & Weissing aseverate in "Animal personalities: consequences for ecology & evolution" (Trends in Ecology & Evolution. Vol 27(8)452-461, 2012, p. 453): "...Behavioral variation is often associated with non-behavioral phenotypic variation (e.g. morphology, physiology & life-history characteristics, or cognitions). Behavioral variation migh often be the cause of correlated non-behavioral variation, given that different behavioral types face different environments exerting differente selection pressures on all aspect of the phenotype. In other case, behavioral variation might be the result of non-behavioral variation, for example if the aggression leven of an individual is made pedendent on the resource-holding potential of the individual or if the activity level of an individual relects its metabolic rate..."

It is possible that there are only well aimed implications alone that your request because the nature of memory, interface of this phenomenon, is not possible to sight. I contribute with orientation bibliography:

MORPHOLOGY & LIFE HISTORY CHARACTERISTICS: BOLNIK DI et al: "The ecology of individual, incidence and implications fo individual specialization" (The american naturalist Vol 161(1)1-28, 2003); BIRO, PA & STAMPS, JA: "Are animal personality traits linked to life-history productivity? (Trends in Ecology & Evolution Vol 23(7):361-368, 2008).

PHYSIOLOGY: RÉALE, D et al: "Personality & emerge of the pace of life syndrome concept at the population level" (Philosophical transactions to the Royal Society (365:4051-4063, 2010); BIRO, PA et al.: "Do consistent individual differences in metabolic rate promote consistent individual differences in behavior?" (Review Cell press, Trends in Ecology & Evolutions, Vol 25(11):653-659, 2010)

I expect satisfy you.

I think yes. Naked DNA doesn’t exist in eukaryotic cells but only on a computer screen or in our imagination. The epigenome roadmap project is proof of it. During its launch by NIH it confronted high profile scientists in the field because of divergence on ideas on fund allocations which resulted in a scientific dispute on the subject.

1/ From a silly point of view, histones are structurally in an Epi-position relative to the DNA even though their tails protrude outside.

2/ They bear a code whose sequence define not only the transcriptional status of the genomic information but also its sub-nuclear location and its 3D conformation. That code also ensures the integrity of the genomic information. In summary, you have to crack the code before you can productively access to the genetic information.

Again, epi-position.

Just to limit my response to the question of whether or not histone acetylation could be regarded as an epigenetic modification, I think yes. Even if the half life of a histone acetyl group is in the range of a few minutes (Histone acetylation/deacetylation) the response is still yes. Imagine (actually it may not be the case) that for a gene to be transcribed it needs a lysine 5/8 H4 acetylation. That requirement will also be imposed to the daughter cell otherwise the transcription of that gene could be impaired (dosage defect) or not taking place at all (null mutation). That gene has an epigenetic memory which might be different from the epigenetic memory of the nearby gene. Memory is a physical not metaphysical entity.

I refer only mental representation that appreciates appetitive or adversative experiences. Of course, memory is a physical entity, but not theirs contents.

To alfonso: I have no idea on how cognitive processes are built up but I recall they were some works showing that HDAC inhibitors have some good effects on lasting memory. I guess because they reestablish a pattern of gene expression that was altered. Anyway, the question posted here is whether or not some short-lived histone modifications could be regarded as epigenetic. I am not sure that in order for a modification to be considered as epigenetic, it has to satisfy the criterion that its artificial alteration has to be transferred to the daughter cell in the absence of the trigger. Still though, histone acetylation does satisfy this condition. Have a look to this work from one of my previous lab (Transient inhibition of histone deacetylation alters the structural and functional imprint at fission yeast centromeres Ekwall K, et al. Cell. 1997 Dec 26;91(7):1021-32).

To Abdelhalim:

You have the reason in your guess that enclosed the activity of histone acetylation in a range of few minues is too an epigenetic modifications subserve for a gene expression or suppression above appetitive & adversative circuitries for a long term behavioral in the context of the confirmation or not the expectative. In these, the recovery for a window time destabilizes the last memoryry to exert the necessary modifications over the contex-adjustment. It is sufficient demostrated that histone deacetylase inhibitors enhance memory process by the activations of key genes regulated by the CREB-CBP transcriptional complex, and histone modifications around individual BDNF gene promoters in prefrontal cortex are associated with extinction of conditioned fear.

Its a lot about convention and definition, but one possibility to discriminate between "real" epigenetic events and all other chromatin-based gene regulation events (that work through posttranslational modification of chromatin components) is dependence on the signal triggering the event: epigenetic modifications are stable even when the signal(s) that triggered the modification, disappear, whereas "conventional" chromatin modifications (non-epigenetic) are dynamic and permanently signal-dependent. This definition would be independent from the "how many generations" question because also a non-dividing differentiated cell would use "epigenetic" mechanisms to stay differentiated, even if there´s no "next generation". But, of course, it would also apply to "epigenetic" inheritance of traits to the next generation (mitotic or meiotic).

Hi Joseph,

I don't agree with you. I think Epigenetics is now a well established discipline. Trying to narrow its field would be a mistake and will generate more confusion than anything else. I don't see how a definition could help in considering the same modification as epigenetic in some contexts and not in others.

House keeping genes will fit well within the scope of your definition while inducible genes will not. Moreover, even DNA methylation in some contexts would be discarded because it is transiient during the activation of the pS2 promoter for example ((Nature. 2008 Mar 6;452(7183):45-50) (Nature. 2008 Mar 6;452(7183):112-5.).

Dear Abdel, as you know the field has been discussing this for a long time. The main problem with "all chromatin&DNA modifications are epigenetic" is, that you would need to include all PTMs of of transcriptional activators/coactivators during normal transcriptional regulation of inducible/repressiible and constitutively transcribed genes. And I dont see the point why DNA methylation should be automatically "epigenetic", when its simply a part of a regulatory cascade during activation/repression and is not transmitted to the next generation. Somewhere you need to draw the line and I agree with those who define epigenetic effetcs as mitotically or meiotically heritable changes in gene expression without changes in DNA sequence.

@Abdel

RE: "I don't see how a definition could help in considering the same modification as epigenetic in some contexts and not in others."

Isn't nature replete with instances where the same or similar mechanisms are used for very different functions?

Dear Joseph and David, please call me Halim,

I understand your point but I resist sharing it for the moment. I think there is no need to include the PMTs of co-activators/co-repressors. We can consider them as readers, writers and interpreters of the PMTs of histones and DNA (or epigenetic code). I think that what you and David are suggesting would be valid only if we find a gene that uses different PMT combinations for the same purpose. But if the same PMTs are seen across different generations then those PMTs conform an epigenetic code independently of whether or not they are transient. A priori, the degeneracy of epigenetic modifications would favor your argument but this degeneracy is only theoretical.

Also, if you read the responses to this thread and the other one posted by David, you will realize that there are people who consider DNA methylation as the only epigenetic modification.

The other thread is at:

https://www.researchgate.net/post/Where_does_the_boundary_lie_between_epigenetic_control_and_non-epigenetic_gene_regulation-particularly_with_respect_to_the_role_of_modified_histones

David, thanks for your statement and for pointing out the other thread on this platform where a very similar discussion is going on. A short comment from my side to the point "heritability of histone PTMs": that obviously not modified histones but the "writer" proteins are inherited is extremely interesting but not decisive for definition of an "epigenetic" event: the outcome - an inherited pattern of gene regulation and transcriptional memory - remains the same.

We know at least one instance of histone modifications that are not transmitted meiotically and that are even probably transient in the cell. H3K4me3 is "written" in spermatocytes at hotspots of crossovers by the PRDM9 protein, used as flags for meiotic recombination, and most probably erased in latter stages of spermiogenesis when vast majority of histones are replaced by protamines.

At least, this only example says that all histone PTMs are not epigenetically transmitted.

Hi Jerome,

What make you say that? Is it because the modification is transient? Is it because of the physical change in the identity of the epigenetic memory carrier due to the replacement of histones by protamines, which are themselves histones-like and they are believed to have evolved from histone H1? They are too subjected to PTM….Or is it because the specification of the hotspots at which the modification happens is somewhat random if we compare them from cell to cell? or ...?

I think the more we discuss this thread the better is the clarity and maybe a better definition will come up. It is true that we are all fed up by the word epigenetic modifications. It is no longer news. A lot of papers only examine a histone PMT to give a boost to their data…etc…

Hi Halim,

Maybe I have not been very clear. It is probably because we don't yet have data to prove unambiguously and directly that indeed, H3K4me3 marks at hotspots are truly erased and-or not replaced by other PTM on protamines. However, we have indirect genetic evidence to say that half hotspots specified in one individual can be different from hotspots in its progeny. Indeed, the PRDM9 writer shows allelic variability. A given PRDM9 allele binds a given set of DNA targets through its variable zinc finger array and writes H3K4me3 on proximal nucleosome(s). For instance, consider a father with PRDM9 alleles A and B and its child with PRDM9 alleles C and B. Both will use B set of DNA targets as hotspots but only father will use tha A set and only child will use the C set. Hotspot usage, which is associated with H3K4me3, is then not transmitted epigenetically here (otherwise child would use A hotspots). I would argue that this is indirect but strong evidence of erasure of the HPTM from one generation to next.

Hope it is not now more confusing than before!!

Hi Gerome! That is really clever to put it this way. I will be thinking about it for sometime. Let see if I understood well. So, hotspots sequence distribution within the genome is predetermined by PRDM9. There are only two possible distributions defined by allelic variants for any individual. What you are referring to is that one of the two possible distributions of potential hotspot sites is different between the father and the son and hence there is no epigenetic inheritance in this case. Half of the sequences bearing H3K4me3 at the level of hotspots will be different between them if I perform PTM analysis. We should also bear in mind, that PTM analysis on hotspots genomic sequence locations will be different between spermatocytes of the same individual.

This is just for my information because you are expert in this field, do you know how many recombination events per chromosome pair happen during meiosis. What it the role of PRMD9 in the mother? Or is it inactivated?

Cheers

Hi Halim,

Half of the sequences bearing H3K4me3 at the level of hotspots will be different between them if I perform PTM analysis.

Yes, with "half" being an approximation: some PRDM9 alleles seem to bind more targets than others. Then one allele can contribute to more crossover hotspots than the other allele in a heterozygote. This is "unbalanced co-dominance" between PRDM9 alleles for hotspot specification. (see discussion in (1))

We should also bear in mind, that PTM analysis on hotspots genomic sequence locations will be different between spermatocytes of the same individual.

That we don't know and this is an interesting question. At the moment, we have an average view (ChIP Seq from millions of spermatocytes) of the H3K4me3 landscape and of DMC1 bound recombination hotspots in mouse spermatocytes (2). There are several thousands of sites in the genome enriched with H3K4me3 dependent PRDM9 peaks and of DMC1 recombination hotspots. Big overlap. One may be able in close future to analyze H3K4me3 landscape from hundreds (tens?) of meiotic cells and see whether there are inter-cell variations.

do you know how many recombination events per chromosome pair happen during meiosis.

At least one, rather two for metacentric chromosomes, meiotic crossover events occur between chromosome homologs. The RNF212 protein defines where crossovers will occur by binding to just one or two recombination sites per chromosome where it triggers the accumulation of the recombination machinery (3). But among these 25-30 crossover sites, approx ten times more were double stand break (DSB) sites, bound by DMC1 in each human spermatocytes. This has been revealed by immunocytology analyses of RNF12 and DMC1 foci on spermatocytes spreads, using fluorescent antibodies against these proteins. Then we know that, for DMC1 there is indeed intercell variation for DSB sites, as ChIP Seq analyses reveals thousands of sites and as single immunocytology analysis reveals only 250-300 sites. It also says (plus many other evidence) that 90% of meiotic DSBs formed are not repaired as crossovers. "Non crossover" events (also coined conversion events) and repair using sister chromatid then represent most recombination events.

PRDM9 ChIP Seq and immunocytology analyses with anti PRDM9 would tell us whether we see the same intercell variation as for DMC1.

What it the role of PRMD9 in the mother? Or is it inactivated?

We have only genetic based data but we can say that hotspot usage also varies between mothers and that this usage is highly correlated with PRDM9 allelic variations (1). So it may well work similarly in male and in female. Again, those who will manage to perform PRDM9 ChIP Seq and immunocytology analyses in oocytes will get very valuable information to fully answer your question.

(1) http://www.sciencemag.org/content/327/5967/836.short

(2) http://www.nature.com/nature/journal/v472/n7343/full/nature09869.html

(3) http://www.nature.com/ng/journal/vaop/ncurrent/full/ng.2541.html

Jerome

I've found these papers (full) very interesting for those who want to deepen the details of the transgenerational epigenetic effects on mice regarding CpG methylation.

Here for the A^vy distinguished agouti phenotypes (only maternal transmission)..

http://www.pnas.org/content/103/46/17071.full

..And here for the Axin^Fu inheritance (both maternal and paternal transmission)..

http://www.pnas.org/content/100/5/2538.full

Both studies report epigenetic inheritances that don't go farther than a couple of generations..

I'm currently investigating the weight of the role of RNA in the inheritance of epigenetic patterns, as that seems to be the case in fairly recent works.

What about this definition by Adrian Bird? ”the structural adaptation of chromosomal regions so as to register, signal or perpetuate altered activity states”. Nature. 2007 May 24;447 (7143):396-8.

Personally I find it outstanding and more comprehensive.