[mxwsn]

The impact of epigenetics can be powerful.

A neat example is sex, one of the most obvious and striking axes of biological variation. Unlike humans where the presence or absence of an entire chromosome (X or Y) determines sex, there are turtles and reptiles that determine their sex based on ambient temperature alone.

The logical consequence is that a pair of female and male turtle twins can share exactly the same genome yet be different sexes — their sex is determined solely by how their genome is read (epigenetics).

[azeotropic]

Epigenetics is not "how the genome is read" -- you're describing the normal action of transcription factors.

Epigenetics is heritable information that is not encoded in DNA base sequence. Common mechanisms include DNA methylation patterns, histone binding patterns, and histone modifications, all of which can be stably inherited from one cell division to the next, but not necessarily transgenerationally, from parent to offspring, through the germ line.

You are also confused about turtle sex determination. The master regulator of sex determination (which must respond directly to temperature) is unknown. The epigenetic regulator KDM6B is required to mediate the temperature signal for male development, but there is no evidence that it is the master regulator. Maybe you read a press release instead of the actual paper in Science.

Environmental sex determination is not that unusual, and need not depend on epigenetics or even transcription factors.

[mxwsn]

"Epigenetics is heritable information that is not encoded in DNA base sequence." I agree with this. Combined with the observation that temperature-dependent sex determination occurs within a specific window of time, after which sex cannot be changed for the remainder of the organism's lifetime, I believe it follows logically that somatic cells (certainly those within the sexual organs) must be receiving heritable information not encoded in DNA base sequence from their adult stem cells. Thus environmental sex determination must occur through epigenetics as you have defined.

To the best of my knowledge, DNA methylation patterns, histone binding patterns, and histone modifications all have causal biological impacts solely by their causal influence on "how the genome is read". Epigenetics is "how the genome is read", as well as more beyond that. But how the genome is read is substantial portion of epigenetics and, for better or worse, the aspect of epigenetics that receives the most attention when discussed with general audiences. I think it's inaccurate to deny the role of "various readings" of the genome in a definition of epigenetics.

I'm unsure how your citing of KDM6B is related to or addresses any confusion in my statement on temperature-dependent sex determination. From what I can understand, you point out that KDM6B histone demethylase acts epigenetically and that it regulates temperature-dependent sex-determination, both of which are consistent with what I said earlier...

[toasterlovin]

> Epigenetics is "how the genome is read"

I'm not a molecular biologist, so take this all with a grain of salt. There are two things which get confused when discussing epigenetics. First, you have gene expression which is mediated by the environment. That is what you are talking about with your turtle example. This is pretty common. After all, being able to respond to environmental stimuli is useful.

But there is another phenomena that gets labelled as epigenetics: stuff in the environment which modifies the DNA of the germ line cells (eggs and sperm) of an organism. The part about it being germ line cells is important, because that means that the changes to DNA will be passed on to the offspring, whereas epigenetic changes in any of the other cells of the body will not. People are in love with this aspect of epigenetics because, frankly, there's a certain type of person who is enamored with the idea that this might "disprove" Darwinian evolution (OMG Lamarck was right!).

It's extremely frustrating because this second phenomena is, as far as we know, extremely uncommon. I mean, it almost has to be. Large organisms are like a house of cards; if you make too many changes, everything comes crashing down. So, if there were epigenetic changes to germ line DNA happening all the time, there would be unfeasible levels of mutation happening from one generation to the next. But the fact that it is extremely uncommon does not stop people from hearing about how this second meaning of epigenetics is a weird end-run around Darwinian evolution, then hearing about how common the first meaning of epigenetics is, and coming to the conclusion that Darwinian evolution is invalid.

Incidentally, these two meanings are the root cause of you and the person who responded to you misunderstanding each other.

[azeotropic]

The somatic gonad does not have adult stem cells (at least in vertebrates). The cells differentiate in the embryo and just stay differentiated. Nothing really needs to be inherited.

Epigenetics is not "how the genome is read" in any meaningful way. Epigenetics is more like encoding tone of voice transcription factors ought to use when they read the genome. Should they whisper or shout? Epigenetics does not do the actual reading of the genome, or change the underlying data in the base sequence. There are not "various readings" of the genome, no matter what you think.

So you haven't read the paper then. Perhaps you also don't understand the difference between mediates and regulates. KDM6B does not regulate the response to temperature. Temperature regulates something (unknown) that regulates KDM6B, which epigenetically regulates something else (in this case, DMRT1, a transcription factor) which regulates a bunch of other things (mostly unknown) that cause the gonadal soma to become testes or ovaries.

That epigenetics should play a bit part in some developmental process like this is essentially a trivial observation, and treating it as significant betrays a lack of understanding. All the genes in all the cells are wrapped around histones which are modified to make the underlying genes more or less active, all the time. So epigenetics a part of gene regulation, but it's hardly the most powerful or important one. It's like waxing poetic over the importance of uridine metabolism, because without uridine there is no RNA. It's true, but in a totally boring way.

As far as transgenerational epigenetics goes, essentially all of the literature in mammals is missing the key controls to rule out regular genetics. You should regard everything you read about it with extreme skepticism. It's a hot field with a lot of hype and pretty lax standards.

[learned]

There is also some really interesting epigenetics research focusing on age-related disease.

For example, Steve Horvath's group at UCLA has been refining an "epigenetic clock" using DNA methylation data to predict all-cause mortality in several species with a relatively simple test. It has the potential to be an incredibly valuable metric in the field going forward.