Name: Christine Mayr
Position: Postdoc, Whitehead Institute for Biomedical Research, Massachusetts Institute of Technology
Background: Postdoc, Institute of Human Genetics, Technical University of Munich — 2004-2005
Postdoc, Ludwig Maximilians University — 2001-2004
Resident, hematology/oncology, Ludwig Maximilians University — 2000-2001
MD, Free University, Berlin — 2000
This week, Whitehead Institute researcher David Bartel and colleagues published in Science findings that chromosomal translocations that have been associated with human cancer disrupt expression of a gene known as High Mobility Group A2, or Hmga2, by the let-7 microRNA.
The disrupted repression, the researchers wrote, “promotes anchorage-independent growth, a characteristic of oncogenic transformation. Thus, losing miRNA-directed repression of an oncogene provides a mechanism for tumorigenesis, and disrupting a single miRNA-target interaction can produce an observable phenotype in mammalian cells.”
This week, RNAi News spoke with Christine Mayr, a postdoc in Bartel’s lab and lead author of the paper, about the findings.
Let’s start with some background on you. As far as microRNAs are concerned, the Bartel lab is about as good as it gets. Did you start off interested in microRNAs before you were looking to do your postdoc?
I’m an MD and I was working in immunology and cytogenetics [in Germany]. Then I heard about microRNAs and thought, “That’s really cool,” and applied to Dave’s lab. I was never working with RNAi before, but I thought that [miRNAs are] one of the major discoveries in the last couple of years and was really interested in them.
Were you ever planning on practicing medicine?
No. I have been working in the lab for several years, and before that I was working in the clinic as a resident, but I will never go back. I’ll just stay in science and basic research. I’m not really interested in the [clinical] applications [of medical research] because I think a major problem MDs have is that they see something and immediately think about treating patients. I think you should first find out a little bit more.
So the Science paper. Can you talk a bit about what led up to the work?
When I started, it was only known that if you over-express a microRNA or if you de-regulate a whole microRNA, it can lead to the fast onset of cancer. There are so many [miRNA] targets, and nobody knows which one of those 200 or 500 targets is really important. [Further, people don’t know] if you need 10 of those targets to de-regulate [the miRNA] or what. Therefore, Dave and I thought that we’d want a more specific approach and really find one single [important miRNA] target.
I was looking in TargetScan — the program that was developed in Dave’s lab to find [predicted miRNA] targets — and I saw [the gene] Hgma2. And this is really the target with the most microRNA sites I’ve ever seen — it has seven let-7 sites that are conserved, and this is pretty exceptional.
I don’t even know if this is a good example for further research because it is so exceptional, but I thought that if I really wanted to show a phenotype with the de-regulation of one single microRNA and target gene, perhaps we should start there.
Can you give an overview of the experiments and what you found?
It was also important that this gene is involved in a translocation that is involved in many human cancers. So we knew already that [loss of] part of the open reading frame and the 3’ UTR … of this gene leads to over-expression of the protein. I thought that perhaps loss of the microRNA regulatory elements is responsible.
The experiments I did were to introduce point mutations in … each of the seven … let-7 sites in Hgma2. Then I did a luciferase assay and could demonstrate that the more sites I have, the more protein I get. So if let-7 can no longer bind to the 3’ UTR of this gene, [the result is protein] over-expression.
And this was dependent on the number of sites I had; with the wild-type of seven sites, I got a lot of repression, and with two sites or four sites or seven sites mutated, I got over-expression. It was really very nicely dependent on the number of sites. I thought, “Wow. That’s cool. We know that let-7 really regulates those sites.” [Prior to this], it was only shown computationally. But I moved on because … I really wanted to show function.
[Hgma2] is an oncogene, and [oncogene over-expression] should lead to cancer, but there are not so many good assays for this. I chose a very old-fashioned assay called a colony formation assay, which I think was developed 40 years ago.
A normal cell requires neighboring cells to grow, but a cancer cell can grow anywhere. So we put … cells into a semi-solid media, and … [transfected the cells with a vector expressing] my constructs — the ones with the mutations in the 3’ UTR where let-7 cannot bind any longer — and [with vectors expressing] wild-type [Hmga2], and some other controls. I saw that I got colonies [with] the constructs where I had the mutated let-7 sites.
Where does that lead you now as far as further experimentation?
In Dave’s lab, it is usually the case where we do one project … and then we do something very different. So now I’m looking for a completely different project in the field of microRNAs. Usually we are not following up [on earlier work] because he always wants to find very new things. Probably [the lab] will follow up on this project and do [experiments in a mouse] … to knock out those seven sites. But this is an experiment that takes some years, but someone will do this in our lab.
What about in general?
We have found two things. One is a new mechanism that can lead to cancer formation. This is when a translocation happens in a human tumor and you get rid of the 3’ UTR with the microRNA regulatory sites, which can lead to cancer. I think this is pretty important because this was not known.
The other thing that was really nice was that, in mammals, the de-regulation of a single gene of a microRNA can be sufficient to produce a phenotype. So far with [miRNAs’] hundreds of targets, nobody had a clue about how many targets had to be de-regulated. It’s still not clear, but we know that in this case one target is sufficient to do something, and I think that’s pretty cool.
Does that finding lend itself to the thought that maybe microRNAs are more reasonable targets for therapeutic intervention than previously thought? If you maybe only need to regulate one target …
This is not clear yet. In the last year, people thought that microRNAs only do a little bit — fine-tuning, but not regulation of a hundred-fold or so. Perhaps in development, but not in the adult cell. Everybody was a little bit worried, but now I think if you find the right gene, you really can see something [significant].
With respect to therapies, I’m probably not the right person to ask because I’m pretty negative on this. I think you have to wait for several years to really find out what’s going on. What we know now is just the tip of the iceberg.
If you introduce a microRNA into a person, what will happen? Nobody has any clue, and I don’t know if you should do this soon. … Many microRNAs are expressed everywhere and they have hundreds of targets. If they don’t do anything, you don’t have a side effect. If they really do something, everything gets out of control.
So it’s a possibility, but we should proceed cautiously.
Let us do a little bit of research, and then ask those questions again.