No kidding. Roberto Ferrari and colleagues at 
UCLA are studying DNA viruses that can change the epigenetic landscape of infected cells to make them more like cancer and less like their old selves. Adenovirus e1a introduces “small e1a oncoprotein” to change fibroblast histone marks — at least, in culture. The virus erases a cell’s previous identity, and apparently alters the cell cycle for repeated cell divisions.
The research offers another window on the crucial RB pathway in cancer. Retinoblastoma, or RB, protein is a key anti-cancer gene with control over cell division, and it’s a main target of this virus. Learning how it affects RB could be important for cancer research, for sure.
I spoke with Roberto Ferrari to learn more about his research into cancer-causing viruses and their epigenetics. He’s lead author on the upcoming Genome Research paper “Reorganization of the Host Epigenome by a Viral Oncogene.”
Virus-induced gene expression isn’t new, but I don’t see a lot of research about virus-induced epigenetic changes. How common is it?
They’re coming up more now. There’s a paper in Nucleic Acids Research — to tell you the truth, I was a reviewer for it — where they actually showed that this other oncoprotein from another virus, which is an RNA virus — a retrovirus — which actually does produce the same modification that we’re looking at.
So, the interesting point that’s coming out right now is that these viruses, in order to cause deregulation of cell cycle — and some of them, to cause cancer, such as papilloma virus — they have to somehow affect two major pathways in the cell. One is fairly well-known, the RB pathway, involving one of the most-mutated genes in cancer.
Another that people are finding more interesting recently is the P300 pathway — P300/CBP is one of the major lysine acetyltransferases. In our Science paper in 2008, we found that you want to specifically affect this histone mark. And a few years later, these other groups came out and found out that if you knock out P300/CBP in vivo, you lose this mark and another mark, which is K27-acetyl.
So what we thought is that these marks can be a very good way to understand the biology of the early epigenetic changes that are happening within the cells, in order to get to a cancer phenotype. And that’s what my boss [Siavash Kurdistani at UCLA] originally found in tumor tissues — that this histone mark, K18, is very decreased. It can be used as a marker of prognosis for cancer recurrence after mastectomy. So the less you have of this mark, the more aggressive your tumor will be.
So the question is, “OK, but what if you use a DNA tumor virus, which is linked to cell transformation — can we see the same thing from the same mark?” And the answer is yes, that’s the case — that’s what we found in that Science paper in 2008. We also found where these markings decreased.
With the advantage of Chip-Seq, we can look at the whole genome, and in this Genome Research paper, we describe what’s happening genome-wide for the first time. The virus is literally erasing all these marks genome-wide, and concentrating these marks to regions that are necessary for progression of the cell cycle. [All this] in a system that would never re-enter the cell cycle. Those [fibroblast] cells are confluent — they’re contact-inhibited.
The virus is creating a system that would allow the cells to reenter the cell cycle. And it does that by sequestering RB, and by affecting RB function and consequently H3K18 acetylation, which would allow the cell cycle to be reactivated again.
When was this first demonstrated, that there was a direct effect on the cell cycle?
In the 2008 Science paper, we used arrays to look at promoter regions of the human genome, and we saw the same thing. There was a global decrease of H3K18ac in promoters in genes related to immune response, but the virus that was increasing the acetylation of K18 at the cell cycle-regulated genes. But at that time, we couldn’t explain the global decrease, because we didn’t know where this mark was before, elsewhere in the genome.
So now, we’ve found that this mark is mostly an enhancer mark — it marks enhancers that regulate genes at long distance. And the virus actually erases the phenotypic identity of the cell. It says, “You’re no longer a fibroblast anymore that expresses collegen genes or whatever — you’re becoming what I want. You have to divide, you have to acetylate cell-cycle-related genes.”
How active is work in this area?
People are finding more and more that there’s a relationship between viral infection and cancer. A couple Science papers a few months ago that showed that in a lot of blood malignancies, P300 and CBP are mutated, and the mutation affects the global level of K18 acetylation.
So that links what is happening normally — a natural mutation of the gene to cause cancer — with what the viruses are doing in the cells, in order to replicate itself. And what it does is very cancer-like.
This is far down the road, but does this point toward any possible cancer treatment options?
It is down the road, but there’s already chemotherapy that includes HDAC inhibitors, which affect the cell’s acetylation state. Now that you know that p300 and CBP mutation can cause cancer, gene therapy, for example, could replace [damaged gene function], and could enhance acetylation of certain sites.
So, other the interesting thing is that most of the K18 acetylation in normal fibroblasts — before infection — is over tumor-suppressor genes, basically. What happens is that this acetylation is definitely linked to more expression of tumor suppressor than oncogenes. When the virus comes in, it pushes this pattern to the opposite direction, and wants more oncogenes to be expressed than tumor suppressor.
So these marks can at least be used as diagnostic tools. To determine, within a tissue, whether there is already development of some sort of malignancy.
We’re far from a comprehensive list of modifications that together might give you a better prognosis or something like that. But we already know that the level of this mark is associated with prognosis of cancer, so there’s something going on.
[The somewhat creepy picture of a television hypnosis program -- or whatever -- is by Flickr user Slavin, and it's used here under a Creative Commons license.]
Ferrari, R., Su, T., Li, B., Bonora, G., Oberai, A., Sasidharan, R., Chan, Y., Berk, A., Pellegrini, M., & Kurdistani, S. (2012). Reorganization of the host epigenome by a viral oncogene Genome Research DOI: 10.1101/gr.132308.111
