It's an interesting study, but the article is somewhat lacking. First off, the title is imprecise. This is specifically about epigenetics (DNA methylation). "Hidden code in DNA" is very vague, and can also apply, for instance, to introns, which are also fast evolving.
Second, it is questionable to what extent the term "evolution" can apply to DNA methylation, since these changes are usually lost within a few generations.
I don't think that it is entirely safe to say epigenetics are inconsequential for Animals. DNA methylation is definitely important to development w/in a single organism, and I think it is still unclear how important it is to heredity.
On a side note:
I am a biochem student w/ an interest in programming.. HN has been really helped me to learn _what_ to learn about computer science. I love when I see good discussion of biology here, but I wish there was a similar online community for molecular biology / genetics. Anyone have any suggestions?
Maybe there is place for an YC startup doing X-hackers news?
Cross-advertisements could get seed users, for bootstrapping the new X variants. (Starting with the present Hacker News, then going to CellHacker News, etc.)
Agree that this article isn't the most descriptive, and evolutionary may not be the best of terms, but it should be noted that epigenetics has caught the eye of developmental and neurobiologists for some time as explaining much more of human traits/behavior than our DNA itself.
The human genome is composed of roughly 20K protein-coding genes, not enough at all to code for all the complexities in our body (if there was a gene for every bifurcation in our capillaries, we'd run out of genes pretty fast; in addition to the fact that there is no single gene for specific human behaviors, i.e. "aggression").
Epigenetics, in its simplest form as DNA methylation, provides a powerful solution to many of the issues of "there aren't enough genes", and explains a lot of the chaos-elements / environmental impacts on heritability of traits.
DNA methylation, by consequence of wrapping the DNA in chromatin, controls access to which genes get turned off/on by transcription factors (TF). TFs in turn are often influenced by environment, especially neonatal. If you do the permutations of which genes are on/off, b/c a TF can/can't access the DNA, b/c the chromatin is on/off = a whole new set of permutations that entail a new information code.
A good example of the impact of epigenetics is obesity. Children's propensity to obesity often has a lot due to with epigenetic influences during neonatal development. For instance, if a mother is pregnant during a famine and thus doesn't eat much, the fetus' digestive system will adapt to starving conditions through epigenetic changes in chromatin that bind the relevant genes, thus developing a "thrifty metabolism" (digests food slowly). This trait then stays through generations, and when such people move to the West, their metabolisms incline them towards obesity as their body still digests food slowly thinking they were born into starving conditions.
So long story shot, epigenetics aren't "evolutionary" in the original sense, but they are redefining how traits/behaviors (too complex to be explained by genes) are inherited/developed.
> A good example of the impact of epigenetics is obesity. Children's propensity to obesity often has a lot due to with epigenetic influences during neonatal development. For instance, if a mother is pregnant during a famine and thus doesn't eat much, the fetus' digestive system will adapt to starving conditions through epigenetic changes in chromatin that bind the relevant genes, thus developing a "thrifty metabolism" (digests food slowly). This trait then stays through generations, and when such people move to the West, their metabolisms incline them towards obesity as their body still digests food slowly thinking they were born into starving conditions.
I think it'd be fair to characterise this as an early hypothesis?
The exact mechanism is still being studied for sure, but the phenomenon is well documented and continually researched (article a couple years back: http://www.pnas.org/content/105/44/17046.short).
The most widely cited example in humans is "Dutch Hunger Winter"
You are confused; traits and behaviours are not too complex to be explained by genes. 20,000 protein-coding genes do indeed code for all the complexities of our bodies, since they can be combined in an essentially unlimited number of ways.
All epigenetics does is determine which genes are turned on at which particular times and at what levels, i.e. it is simply another mechanism relating to the regulation of the protein-coding genes (which itself has been well-understood for decades).
'Epigenetics' is simply a fashionable term that plays well on grant applications. It's not unimportant, but neither is it revolutionary.
Apologies for the lack of clarity. When I said 20,000 protein-coding genes are not enough, I was referring specifically to the traditional view that one gene codes for one protein. It's pretty well understood now that this is not the case, and my point was moreso that 20,000 genes in and of themselves fail to code for every different trait/behavior in a reductive model. I think epigenetics provides the powerful toolbox to illuminating on the unlimited # of interactions/permutations of genes that you describe.
But what I find most fascinating about epigenetics is its ability to explain environmental vs biological effects on our traits/behavior, not intrinsically described by our genes. As the DNA-methylation described in epigenetics determines which genes are on/off, it is external stimuli that often affect the methylation of epigenetics in the first place. (a Nature brief on the topic in mice: http://njc.rockefeller.edu/PDF_BN08/topic%202-FrancisEpigene...) It's a fascinating way of understanding how our environment affects us on a biological level.
I agree the article is somewhat lacking, imprecise, and questionable. Also, it is not clear what is really new here, except maybe the developed a method to study this with plant clones (so they can look at just the epigenetic variations).
I think DNA methylation is just one type of epigenetics -- literally above the genes -- it can refer to any information encoded on top of nucleic acid base pairs.
Would you say that cultural does not "evolve" because the changes are usually lost within a few generations? Is "evolution" a trademark of classical genetic evolution scholars?
Also, at least according to NOVA on PBS in 2007, methylation epigenetics (which vary from cell to cell in the body) can be modified during one's life and influence inheritance. The NOVA program also claims methylation epigenetics accounts for the difference in phenotypes in identical twins. [edit: grammar]
I agree that "evolution" is usually a rather vague concept. In biology, however, we do have Darwin's definition based on the "strong principle of inheritance", involving differences in fitness among hereditary variations. Both methylation and cultural change fail the "hereditary" criterion in this strict definition.
It might be possible to analyze cultural change, to some extent, on these lines, provided the units of selection are well picked. (Dawkins made a brave attempt with "memes", but the only credible uses of that theory that I've seen involved peripheral phenomena, like earworms and various other inconsequential fads.) Methylation doesn't seem to fit that bill.
No wonder that there are people who believe in "intelligent design" with headlines like that. The word 'hidden' could be interpreted as someone or something has hidden it there.
Second, it is questionable to what extent the term "evolution" can apply to DNA methylation, since these changes are usually lost within a few generations.
Third, epigenetics are relatively inconsequential for animals, for which the Weismann barrier is hardly ever penetrated: http://en.wikipedia.org/wiki/Weismann_barrier