Dr. Andrew Feinberg discusses how recent events are reshaping the way his team thinks about DNA methylation in cancer. This interview was shot on campus at Johns Hopkins University.
This video was brought to you by epigenetics supporters at EMD Millipore.
Redefining DNA Methylation in Cancer
Well, there are a number of things I’m excited about right now. Our own research has been transformed a lot because of the advent of whole genome techniques for assaying methylation and other epigenetic marks, some of which we’ve been able to develop in our own center by working closely with our statistical colleagues.
But the highlights for me, what I’m most excited about at the moment are first of all, a re-analysis of DNA methylation in cancer. So obviously we’ve been working on this since the beginning days of studying DNA methylation in cancer. But our ability to look across the whole genome has led to some very surprising findings. So, one of them was the discovery of CPG island shores, which are these regions near the islands themselves of intermediate CG density that seem to show a greater degree of methylation variation, almost regardless of the question, but in cancer in particular, then we see within the islands themselves.
Large-Scale Hypomethylation
The second thing comes from our recent whole genome bisulfite sequencing analysis that was published this past year in Nature Genetics in which we were able to ask, for the first time all the way down to the base level what’s actually happening to DNA methylation across the genome in colorectal cancer. And one of the things that really surprised us in this work were that there were these gigantic regions that we call “blocks” of large-scale hypomethylation across the genome involving about a third of single copy genes.
And I think that might account for much of what Bert Vogelstein and I discovered in 1983 when we first reported ultra methylation in cancer. And what’s really interesting about this, too, is that the genes that lie within these hypomethylated blocks show the greatest degree of hypervariable expression in cancer. But moreover, it converges with another observation that we were able to make a couple of years ago by using some of these high throughput techniques of what we call LOCKs.
These are Large Organized Chromatin and the “K” stands for lysine modifications that appear to increase developmentally and may also show some tissue variation and that also correspond to the lamen – nuclear lamina associated domains that have been described by others. So our hypomethylated blocks turn out to correspond to these very large regions of LOCKs. And that’s very exciting, when you have convergent data in your own laboratory. And so we think that there might be a mechanistic relationship between the developmental plasticity, the ability to allow genes to become expressed and then get restricted expression to development and a release of that process that takes place in cancer. So, that’s very exciting and we’re quite interested in that.
Epigenetics Outside of Cancer
Another area that I think is finally reaching maturity for our lab and a lot of other labs is to really understand in a detailed way what’s going on in the epigenetics of common diseases outside of cancer. So while this has been subject to a lot of review articles over the last few years, now our lab and other labs are very interested in understanding from this sort of new perspective we call epigenetic epidemiology what’s the relationship between individual variation in disease, common disease, in DNA methylation or other types of epigenetic information and genetic variation and how that might be transmitted in families, and how that might be modulated by environmental exposure.
So there are a number of areas that we’re pursuing in that field. One of them is autoimmune disease. Another one is the regulation of glucose metabolism. And I think that will be a fertile area of study going forward.
Developmentally Regulated Stochasticity
And then I guess the third thing that excites me a lot has to do with this idea that I’ve been – become very interested myself in the last few years of developmentally regulated stochasticity. In other words, can variation itself – phenotypic variation itself – be an important property in normal embryonic development and also in the ability to respond to environmental stimuli? And might that be itself regulated developmentally? And how might that process be abrogated or disturbed in some way in disease, including cancer?
I mean, our own data from the colon cancer sequencing – bisulfite sequencing project shows that a mechanism just like that might be very important in the unstable regulation of gene expression in cancer. But another way of pursuing it that we’re following ourselves right now is looking at a model organism. So we’ve actually been studying honeybees.
So this might be surprising to people who know, you know, the work I’ve been doing in human genetics. But with a wonderful collaborator, Grovian Amdan at Arizona State University, we’re asking whether or not developmental changes in the brains of honeybees might account for the heterogeneous pattern of behavior that’s also maturation dependent in naturally occurring populations that are related to a very interesting phenotype, namely behavior. And to what degree that might be reversible. So, those are some of the things that I think are most exciting me at the present.