Today in my class we talked for a while about epigenetics. I used it as an example of a term we’d encountered more than once in our ecological developmental biology course, but that has some complicated ambiguity and fuzziness that has led to all kinds of weird popular confusions about the subject. I was also using it as an example of critical analysis of a paper, as I discussed yesterday, and it was a lead-up to having the students discuss papers on relevant topics they were interested in — so we spent most of our time talking about other things.

But I’m going to talk now about just this one paper I read. You see, Larry Moran and I have been having this long-running disagreement about epigenetics — nothing hostile, just an occasional cocked eyebrow in each other’s direction — which you can see on display in this article by Larry on epigenetics, in which he disagrees with my definition of epigenetics, back in 2008. Here’s my definition:

Epigenetics is the study of heritable traits that are not dependent on the primary sequence of DNA.

And here’s the definition used in Gilbert’s text:

…molecular processes around DNA that regulate genome activity that are independent DNA sequence and are mitotically stable.

And here’s Larry’s objection:

Here’s the problem. If this is epigenetics then what’s the point? When I was growing up we had a perfectly good term for these phenomena—it was regulation of gene expression. Why is there a movement among animal developmental biologists to use “epigenetics” to refer to a well-understood phenomenon?

While I agree that “epigenetics” is a huge, broad, diverse category of phenomena, I think he’s overlooking a key point to claim it is synonymous with gene regulation. It is gene regulation that is heritable and mitotically stable. It’s still far too open-ended, but it’s not just any old example of gene regulation.

It’s also clear and consistent. Larry challenges us with eight instances of regulatory phenomena and asks which ones qualify as epigenetic. Easy. 1, 2, 6, 7, 8. Those are the ones where he specifically mentions multi-generational inheritance of a regulatory state. 3, 4, and 5 describe responses within a single cell in a single generation (5 is sneaky, though: Drosophila oocytes are having gene expression modified in ways that might be transmitted through multiple generations — it’s just that those cells are being loaded with bicoid RNA, not having their bicoid genes being set to a sex-specific state).

I am also comfortable with the idea that inheritance of the regulatory state of the lac operon is an example of epigenetics. It’s arguable whether that’s a useful category, but it does fit the definition.

So one approach that could be taken is to come up with a more specific or more practical definition.

Larry has a more recent article in which he agrees with a new paper by Deans and Maggert that tries to do exactly that. It also takes a much appreciated historical approach, giving the various definitions that have been wafting about since the 1930s. For instance, here’s Waddington’s ancient physiological definition:

the branch of biology that studies the causal interactions between genes and their products which bring the phenotype into being

Yes, I agree — that would simply be gene regulation nowadays. You can’t blame us wicked developmental biologists for promoting that one, though, because we don’t use it anymore.

Now we favor the Holliday definition:

the study of changes in gene function that are mitotically and/or meiotically heritable and that do not entail change in DNA sequence.

To me, “heritable” is the magic word that makes all the difference. This, however, is not enough for Deans and Maggert. They want to add more focus, often a good thing, and narrow the definition. I was not happy with their argument, and thought it poorly made, though. See if you can find what was objectionable in this section of their paper (I highlighted it to make it easy, an epigenetic modification that does not change the sequence of the letters in the text.)

We don’t feel that it is possible to reconcile Waddington’s focus on gene regulation with Holliday’s more specific criteria within one field and still maintain the level of clarity needed to produce a useful definition. The efforts to preserve a relationship between these two conceptualizations have been impaired by the fact that there are just too many phenomena, with too few mechanistic connections, to categorize into one field. Also, among the definitions that do maintain the requirement of heritability, we feel that many lack the detail to be functionally useful in directing the testing of specific hypotheses, particularly as it relates to the location or site (cytoplasm or nucleus) of epigenetic phenomena. To mitigate these shortcomings, we advocate defining epigenetics as “the study of phenomena and mechanisms that cause chromosome-bound, heritable changes to gene expression that are not dependent on changes to DNA sequence.”

We feel that this definition makes a strong distinction between gene regulation (Waddington’s definition) and epigenetic inheritance (Holliday’s definition), and also emphasizes that epigenetic phenomena must deal exclusively with chromosome-bound changes. By making these distinctions, we have efficiently separated expressional changes caused by cytoplasmic compounds, which are more closely tied to gene regulation, from those which occur on, or in close association to, the chromosome. Doing so makes the focus of the field much clearer and identifies epigenetic mechanisms more explicitly.

We feel that this definition touches on several important elements not encompassed by other definitions, yet commonly implied in most uses. To further explain the reasoning behind our definition, as well as its utility for improving epigenetic research, we would like to offer a clarification and a test.

Yeesh. I don’t feel that your personall feelings are a strong argument, and I cringed when I hit that page. At least edit it to remove the emphasis on your personal discomfort; just say that the old definitions lack detail, rather than that you feel they lack detail.

So let’s pull out their shiny new definition.

the study of phenomena and mechanisms that cause chromosome-bound, heritable changes to gene expression that are not dependent on changes to DNA sequence

Well. All this fuss for a single change, the addition of the phrase chromosome-bound. That’s it. I agree, it does narrow the topic, but it’s still covering an awful lot of territory. I’m not feelin’ it. I have the impression that the primary virtue of the new definition is that it reduces a class of phenomena to a subset that many people are comfortable studying already, and in part reinforces a gene-centered perspective on cellular behavior.

It also leaves me wondering…what about the inheritance of cytoplasmic or membrane-bound factors that induce consistent changes in gene expression in daughter cells? The gene regulation aspect may be mundane, but it’s the inheritance that is interesting. Under the Deans and Maggert definition, this is no longer under the umbrella of epigenetics — it’s something different for which we have no general name now.

It makes Larry happier, though.

I think this is a useful definition. Nobody cares if dividing E. coli cells inherit molecules of lac repressor and continue to repress the lac operon. That’s a trivial form of epigenetics that never posed a threat to our understanding of evolution.

That’s odd. I do care that the lac repressor is cytoplasmically inherited, but then my primary interests, in the most general form, would be in the patterns of stability and change in cellular properties, rather than the metabolism of sugar. Telling me that I should only pay attention to inherited proteins or methylation states that are directly bound to DNA seems arbitrary.

I also don’t consider “poses a threat to our understanding of evolution” to be a relevant criterion. I agree that lac repressors don’t challenge evolutionary theory, but neither do heritable histone modifications or methylation. I’m one of those people who think epigenetics (even under the old definition!) is important and interesting, but doesn’t affect evolutionary theory much at all.

Larry and I agree.

Methylation is trivial.

Well then, if inheritance of the lac operon is such a trivial form of epigenetics that it should be excluded from the definition, then we apparently need yet another definition that excludes the triviality of methylation.

Or, really, we should recognize that “trivial” is not a good reason to exclude something.

I will still second Larry’s argument that none of this stuff overthrows modern evolutionary theory in any way. It would require extremely persistent inheritance of an epigenetic state over many generations to have those kinds of repercussions.

(The Gilbert text does mention one significant effect: the toadflax plant, Linaria vulgaris, has a radically different flower morph, Peloria, that Linnaeus himself classified as a different species. As it turns out, they only differ in the methylation state of the cycloidea gene, but the DNA sequence is identical. This is a case of an epigenetic change persisting for hundreds of generations. It’s a rare case, though, and also…would still definitely fall under the Deans and Maggert definition.)