For years, I’ve been telling my students that epigenetics may represent a “Lamarckian” inheritance mechanism. Small chemical modifications to DNA that do not affect the actual sequence (of A, T, G, C) can control gene expression. These markers are acquired during our lifetimes as we experience and interact with the world. Some evidence has suggested that these markers are heritable–able to be passed down from parent to offspring.

Recent studies in epigenetics (those chemical modifiers of DNA) are being heralded as the first evidence that these changes may represent a force for evolution. This blog post presents an analysis of the recent article by Greer et.al. on the epigenetics of Caenorhabditis elegans and how it relates to evolution–and how strong I think the evidence is.

You can read a more general account of the story at http://www.nature.com/news/2011/111019/full/news.2011.602.html

First, let’s understand in general terms what the authors were researching.  They were interested in learning whether a particular epigenetic modification could be inherited–and extend lifespan.  They used the model organism Caenorhabditis elegans.

The researchers were looking at the cellular machinery (enzymes) that make a modification to histone protein #3.  Histone proteins help organise DNA in a cell: the strand of DNA wraps around the protein to help condense it into a chromosome.  This is how you fit a 2.8-inch piece of DNA into a microscopic cell.  Although it’s easy to imagine histones as spools of (DNA) thread, that’s not exactly a complete picture.  Chemical modifications to these spools can regulate gene expression: which genes are “on” and which are “off.”  In their studies, Greer and his colleagues were looking at modification of histone H3.  This particular modification was the addition of three methyl groups (-CH₃) to the fourth lysine (amino acid) of the protein.

This modification typically activates genes (turns “on”), which was reported in another Nature article. The “machinery” is actually a group of three enzymes that work together (collectively termed a holoenzyme or complex).  The three parts will be important, so let’s look at what they do:

1. ASH-2: transfers the methyl groups and shortens lifespan

 

1. WDR-5: regulates ASH-2 and shortens lifespan

 

1. SET-2: checks nearby lysine 9 for methyl groups.  If absent, will add methyl groups to lysine 4, which activates genes.

Based on what we know, a defect in any of those three parts should make the whole thing stop working, which means genes can’t get that three-methyl “on” modification: the target genes should stay off and lifespan should increase.

What Greer’s group did next was to breed a lot of C. elegans.  In particular, they bred normal worms (+/+, wildtype) to WDR-5 mutants (wdr-5/wdr-5, knockout).  This generated heterozygous (+/wdr-5) worms that were allowed to reproduce for a few more generations.  Interestingly, both the heterozygous worms and the knockout worms lived about 20% longer than wildtype worms!  Hey–that matched our prediction hypothesis above!  By the F5 generation, though, (that’s 5 generations after our original homozygous parents), only the homozygous knockouts (wdr-5/wdr-5) still showed longer lifespan.

So–did it work for Set-2?  Yes.  They got almost exactly the same results, but saw a 30% increase in lifespan for generations F1-F4.

What about Ash-2?  Yes.  They had to use a different technique–but they still got about 20% longer lifespan for 4 generations.  However, the F5 generation showed no difference.

Remember where I put emphasis on heterozygous above?  Well, those worms have one good copy of the gene and one bad copy of the gene (+/wdr-5). That means their good copy should still be able to work–and turn ON genes–and keep lifespan SHORT.

But it didn’t work that way!  This shows that the pattern that was already in place from mommy knockout worm was somehow passed down–INHERITED.  Ok, so mommy was a hermaphrodite, but, well, whatever.  No family is perfect.  This happened for all 3 genes.

Could it be that the worms needed both copies of the gene for everything to work right?  Like one copy wasn’t enough to fix all the problems?  Maybe.  If that were true, then heterozygous worms (+/wdr-5) should have lifespans between the normal and knockout worms–but they didn’t; their lifespan was the same as the knockouts, which means no “repairs” happened.  They even did additional experiments to demonstrate that the “normal” copy was working properly (Western blot) and showed that histone 3 (H3) was in fact getting methylated like it was supposed to in the wildtype and heterozygous worms (but not the knockouts).

And that’s the really cool part.  See, those methyl groups were thought to be removed between each generation, and this research showed that they’re not!  They can be inherited!  Epigenetic research is still very, very new, but we’ve been able to demonstrate that

“…epigenetic inheritance has been described for some traits, including flower symmetry and colour in plants^1–3, progeny production in worms⁴, heat stress response and eye colour in Drosophila^5–7, and coat colour in mammals^8–10.”

 

 Darwin’s Postulates Regarding Evolution

 

1. There must be variation in the population: in this case, yes.  Some worms lived longer than others.

 

1. That variation must be heritable: Again, yes.  This research showed that these differences in lifespan could be passed through approximately four generations.

 

1. Survival and reproduction should be non-random and tied to that heritable variation: in this case, we have a PERFECT example; the genes themselves control lifespan directly, so yes, of course survival was different!

What made this research so unique and important is that is shows us a new mechanism for inheritance.  Not all of our traits are coded by A, T, G, C (directly in the DNA)!