At A Glance
Name: Erik Sontheimer
Position: Assistant professor, biochemistry, molecular biology, and cell biology, Northwestern University
Background: Postdoc, University of Chicago — 1993-1999; PhD, molecular biophysics and biochemistry, Yale University — 1993; BS, molecular biology, Penn State — 1987
While Erik Sontheimer had been watching the RNAi field develop for some time, his involvement with the gene silencing technology really began when Richard Carthew joined the faculty of Northwestern a few years ago. The two have been collaborating ever since, but Sontheimer’s RNAi work has also taken on a life of its own.
Recently, Sontheimer spoke with RNAi News about his efforts and the RNAi field.
How did you get started with RNAi?
My training and most of my background was in the area of pre-messenger RNA splicing, and I have a general interest in gene expression, in particular the roles RNAs play in gene expression. Therefore, even before I was working on RNAi, I had a basic interest in it — I followed the field and it was a very exciting field, and it was clearly something very new and unanticipated.
I watched with interest as it developed in 1999 and 2000. Then, in the summer of 2001, Rich Carthew moved to Northwestern [from] … the University of Pittsburgh. He was the one who, in 1998, first demonstrated that RNAi functions in Drosophila. This was a few months after the original report by Andy Fire’s group and Craig Mello’s group of RNAi in C. elegans.
So [Carthew] has a long-standing interest in RNAi and he moved to my department, and we started talking at various times. He mentioned to me that they we setting up a genetic screen to try to identify RNAi-defective mutants in Drosophila. So, his interest in it was primarily genetic — he has biochemistry in his background as well.
[Carthew’s genetic screen] raised a very interesting possibility, and that was that flies would provide the perfect system in which to try to combine genetics and biochemistry. It has already been shown by Phil Zamore, Phil Sharp, Tom Tuschl, Greg Hannon, and so on, that Drosophila cells from various sources can be a good source of in vitro RNAi activity, and those groups were working on it pretty hard and were making rapid progress.
But nobody had done a good genetic screen in Drosophila. There had been good genetics done in other organisms such as C. elegans [and] plants, but those organisms don’t really offer the same opportunities for biochemical analysis. So, based on the work that was initiated in the Carthew lab, we starting talking about the possibility of collaborating on this because my lab worked at the biochemical level on RNA and gene expression, and therefore we had some ideas about how to do things.
The possibility here was that this would open up the door to trying to make in vitro extracts from the mutants that were being isolated in the Carthew lab. If you can combine genetics and biochemistry that way, that can really accelerate progress because, in principle, you can define the mechanistic basis for the RNAi defect in any given mutant.
Most of this was prompted by Rich, and then we came on board while the screen was already underway and we started testing some of his mutants. In the process of testing his mutants, it occurred to us that we needed a better way, a more streamlined way, of trying to look at the complexes that mediate RNAi.
From my background in the splicing field I was well aware that doing native gel electrophoresis had been a very powerful technique, so we decided to try that. It worked quite nicely and that led us off into a number of different directions — it took us beyond the initial scope of the collaboration.
Can you touch on those different directions?
Sure. The genetics were absolutely critical in all of this because one way of validating the different complexes we saw was to find out that they were, in fact, defective in the presence of certain mutations that block RNAi. That means that the complexes we’re looking at are likely to be real — they’re likely to be true RNAi complexes and not some unrelated RNA-binding activity.
Once we were able to visualize these complexes and assay for them, that gave us a biochemical handle on them that allowed us to characterize them further and it eventually allowed us to gain some further insight into the role the Dicer-2 protein plays in RNAi and also to fit all these complexes into an overall assembly pathway. Finally, just from characterizing these complexes, it became clear that one of the complexes that we saw had all the properties that one would expect of RISC and it was a much larger version of RISC than anybody had seen or reported up until now.
It moves beyond just simply trying to identify the defect in any given mutant. It was more of a basic approach to analyzing RNAi-related complexes and trying to define an assembly pathway for a functional RISC complex.
What other sorts of projects do you have ongoing or are you looking into?
We have a lot more work to do on these complexes that we have identified, so we’re certainly pursuing them. And there are still plenty of questions that remain about the overall assembly pathway and we’re looking at that.
We have some projects going on [that explore] the biochemical mechanism of RNAi in human cell extracts.
What sort of experiments?
Well. It’s a very, very competitive field and we’re at a relatively early stage. I think if we just leave it as a generic statement of “the biochemical mechanisms of RNAi in humans,” we can leave it at that.
You talked about issues related to the assembly pathway? Could you break those out?
What we observed were three different complexes. There’s reason to believe that one of them, that we refer to as R1, is similar or identical to a complex that had been identified in Xiaodong Wang’s lab at the University of Texas Southwestern Medical Center, and that is a heterodimer of Dicer-2 and another protein called R2D2. They had found that that complex is able to bind siRNAs and, furthermore, that R2D2 was essential for the incorporation of those siRNAs into a functional RISC complex.
What we have found is that it’s not actually so much that siRNAs get passed from one complex to another, but rather where the entire Dicer-2/R2D2/siRNA complex gets assembled [they] get incorporated into a functional RISC complex. So, it’s not so much a handoff mechanism as it is an assembly mechanism.
We also observed another complex that we refer to as R2 and that one we know far less about. There’s still a lot of work that we have to do in order to characterize it. It’s a little harder to work with — it doesn’t seem to survive biochemical fractionation as well. In fact, it’s still formally possible — we don’t believe this is the case but it is formally possible — that it might even not be on the pathway towards the functional RISC complex. We can’t really exclude that at this point.
Finally, the RISC that we see — we refer to it as the R3 complex — is huge. Something that sediments in a glycerol gradient or sucrose gradient at 80F — that makes it the size of the eukaryotic ribosome, roughly, and that’s somewhere between 4 and 5 megadaltons. That is a very, very large complex, and it may well be associated with ribosomes. We would like to determine whether that’s the case. [We’d also like to know] in general, what are the components of this very large complex?
Do you have any thoughts on where this is heading … in five years or so?
One thing that has been interesting about this field is that it started out in a relatively focused fashion where double-stranded RNA inhibits gene expression. Then it quickly was discovered that it does so by degrading messenger RNA. At the time, it just seemed like that might be the extent of it, but one thing about RNAi is that it has continued to expand in terms of our realization of what aspects of biology might be controlled by it.
Now, it’s not just a question of messenger RNA degradation, it’s also transcription, chromatin, translation. It’s all these different areas of biology. Since we’re still in the middle of that — it’s less than half a year ago, I think, that people discovered a complex that uses small RNAs to form heterochromatin — it’s kind of hard to say where the extent of the importance is going to finally stop.
Once that happens, it becomes a question of getting to greater and greater levels of details, greater and greater levels of understanding, of how the process actually works. That will take us well beyond five years.
The splicing field … was first discovered in 1977, the first in vitro systems established in 82 or 83, and there’s still a lot of people working on it — getting to where you completely understand the system takes a lot of time, so I think there will still be a lot of activity on the basic mechanism in five years.
As for the therapeutic side, that’s harder to predict. There are a lot of people that are concerned about, in particular, delivery — how to get an siRNA to the right place without first getting degraded. That’s a tough nut to crack.
That’s a field that I’m not an expert in so it’s hard for me to make predictions on that.
It’s a bit shocking that there was this mechanism with a role in so many different areas and that people didn’t have any idea about it for so long.
That’s right. [RNAi] really flew under the radar for a long time and it’s really amazing the way it’s completely exploded.
Phil Sharp has made the point that there’s really nothing that prevented us from recognizing it 20 years ago. The discovery of RNAi did not require any kind of highfalutin, new, cutting-edge technology that didn’t exist back then.
What was it then?
There’s a common problem of ribonucleases [where] enzymes just chew up RNA in a very non-specific fashion. That’s a perennial problem for people who work on RNA. I think most people had the sense that if you go looking for RNAs that are that small, all you’re going to get is garbage — things that were normally a lot bigger but got degraded non-specifically.
A common technique is gel electrophoresis and that part of the gel was pretty much always run off the bottom; you never even looked at it. I think that was really the main thing and it was a great realization on the part of Andy Fire and Craig Mello and the others who really pioneered the field that … these things really should be looked at and are important.
It also coincided with microRNAs. The first microRNA was discovered in 1992 or 1993 by Victor Ambros and Gary Ruvkun, and it just seemed like such an oddball. It was just baffling to everybody how this tiny RNA could have anything to do with anything. They were just way ahead of their time.