The National Institutes of Health (NIH) have launched a $190 million research effort to learn about epigenetics. According to Nature News ...
It’s not that epigenetics is totally useless. I just don’t see why it’s worth 190 million dollars.
Kevin Struhl
Harvard Medical School
Epigenetics, described as "inheritance, but not as we know it"1, is now a blisteringly hot field. It is concerned with changes in gene expression that are typically inherited, but not caused by changes in gene sequence. In theory, epigenetic studies can help explain how the millions of cells in the human body can carry identical DNA but form completely different cell types, and perhaps why certain cells are susceptible to disease.It seems strange to be spending so much research money on something that scientists can't even define properly [Epigenetics in New Scientist, Epigenetics Revisited, Epigenetics]. Perhaps some of this money will go toward coming up with a reasonable definition of what they're studying.
The NIH's epigenomics initiative is a plan for such studies on a grand scale including, for example, surveys in different human cell types of all the chemical tags, or epigenetic marks, that might control genes.
As far as I know we already have a good model for how totipotent cells can differentiate into many different cell types. We also understand how they can revert to pleuripotent cells. We even know that differentiated cells can be transformed back into stem cells by adding certain transcription factors.
Now we add in the fact that some DNA binding proteins can covalently modify DNA and this affects gene expression. What's the big deal?
Presumably one of the big deals is in understanding the manner in which epigenetic chromatin modifications are regulated in a gene-specific manner.
ReplyDeleteThis really makes no sense at all. The papers that have come out last year mapping chromatin modifications in ESCs, NPCs, MEFs, and CD4+ cells have already shown what we can expect to see - H3K4me3 enrichment around TSS, nucleosome free regions, bivalent chromatin domains that get resolved in some cell-specific way, and so on and so on. Some of these will correlate with expression or repression, others will not, and that's it.
ReplyDeleteBut how these marks are set up will not be answered. The only certain benefit will be the prediction of some new genes that have escaped detection so far.
Mapping all TF binding sites (and not in the static and purely descriptive manner it has mostly been done so far) would be a much better use of $190 million dollars. The big challenge is understanding development in detail and being able to predict its outcome, not doing librarian work.
We should definitely have a catalog of chromatin marks in all cell types of the body, but it is much wiser to wait for sequencing to become cheap enough for such a project to be cost effective, if we have a rough idea what we're going to get out of it, waiting 3 or 4 years isn't a big deal
But how these marks are set up will not be answered.
ReplyDeleteand you know this because... ?
Adrian Bird has an excellent article about the meaning of epigenetics in Nature. Quick and dirty is that most working in the field define it as heritable chromatin modifications and DNA methylation. Which is funny for those of us working in neurons, as they don't divide and therefore technically can't have heritability.
ReplyDeletehttp://www.nature.com/nature/journal/v447/n7143/full/nature05913.html
The way epigenetics seems to be defined, and its consequent pervasiveness, makes it totally worth $190 million. Pushing stem cells down a certain developmentary path - epigenetic. Learning and memory - requires histone modifications, requires gene-environment interactions - epigenetic. Cancer - huge epigenetic modifications. Effects of early environment on gene expression - involvement of epigenetics. Immune system memory - epigenetic.
Mechanism is of tremendous interest to some of these researchers. If we want to see how the marks are set up, it's going to cost money, and take time. $190M over 10 years is a good first step.