Evolution links

Essay about the controversy in Science
Nasty Dawkins diatribe
Examples of coopertivity
Wiki on group selection
Pinker Essay
DS Wilson article for counterpoint to Pinker
E.O. Wilson paper from 2005 that starts this mess


Epigenetics. Or, how do you change DNA without changing DNA?

OK. That was too clever. What I mean is, how do cells permanently become determined to do only a subset of things without changing the DNA sequences?
We know that Immune cells actually do change some of their DNA sequence. But, that's really unusual. When epithelial cell divides, the two daughter cells "know" they are epithelial cells. This comes down to changes in how the DNA is packed and arranged in the nucleus. Whole regions of chromosomes are tightly packed in what we call "heterochromatin," which is differentiated from open "euchromatin," and is not actively transcribed.
Moreover, some of the changes are more subtle than that, but still maintained as the cells divide. So, the questions are:
  1. what are these stable "marks" on the DNA?
  2. How are they maintained?
  3. How are they removed?

That last one is really important. Because we know how that works, we now can convert adult cells into "Induced Pluripotent Stem Cells," IPSCs, which can do any job.

The answer to the first question basically is:
  1. there are changes made to the non-base-pairing sections of the DNA bases, usually addition of methyl groups at Cs in "CpG" islands
  2. additions of methyl groups to the proteins that pack the DNA (histones)
  3. there are added acetyl and phosphate groups added to the proteins, which tend to "open up" the DNA.

  1. The first two generally close down the DNA and make it less accessible to transcription.


These are maintained by enzymes such as these:

which read the status of the DNA after replication and re-establish it. So, if these enzymes find that one strand of DNA has methyl groups on the 5'-C-p-G sequence (that would be the template or "old" strand), they put methyl groups on the other stand (which would also be 5'-C-p-G)


The answer to the last question is that there are transcription factors that, when induced, will turn on a battery of genes that "scrub" the marks off all the DNA and histones, essentially rendering the cells back to the default, pluripotent state. How would you go looking for them? As we discussed, the experiment that worked was to transfer genes from embryonic stem (ES) cells to ordinary cells and see if you can find genes that induce the ordinary cells to become more ES like. The method here is a bit more difficult than the "transfer of genetic material" expeiment we did. But, the idea is the same. They used a viral system to "transduce" genes. Originally, 8 different genes were found to be needed. Since then, it seems you can pare down that number to 2 or 3 genes (which often then induce the other 5 or 6).
After that, scientists did experiments in which they knocked out one or more of the genes they determined were needed to either induce or maintain the "mark-free" state, often finding that th embryos would differentiate prematurely into a "trophoblast." Basically, these are the first differentiated cells in an embryo and are technically never going to be part of the developing organism. they form the support cells around the embryo and, in placental mammals, form the embryo side of the placenta.
This result adds support for the idea that the genes are involved in establishing or maintaining the "pluripotent" state.
You might like the relatively short wikipedia article on IPSCs, which is found here.