At A Glance
Name: David Sabatini
Position: Associate member, Whitehead Institute/ associate professor of biology, MIT
Background: BS, biology, Brown University — 1990; MD/PhD, neuroscience, Johns Hopkins University School of Medicine — 1997; Whitehead Fellow, Whitehead Institute — 1997-2002.
Although he received a medical degree from Johns Hopkins, David Sabatini decided against entering a residency program and pursuing a career in medicine. He instead joined the Whitehead Institute’s fellows program so he could conduct independent research, and ultimately landed a faculty position at the institute.
Recently, Sabatini took time from his teaching and research duties to talk to RNAi News.
How did you first get into RNAi?
The first time I was introduced to it was when Jerry Fink, whose was the director of Whitehead at the time, gave what we call a Whitehead Forum — this is basically when someone gets up in front of a large part of the institute for some kind of scientific talk. He had thought that there was a trend he had noticed in a couple of organisms of this sort of RNA interference mechanism; he gave a talk about it and it really kind of woke up people about [the technology].
[This was] before … mammalian RNAi had come out, it was before Dicer, it was before 21-mer siRNAs — [the talk] was an appreciation that if you put double-stranded RNA into certain organisms, there was a silencing. It was before a lot of the things we take for granted now.
After you heard the talk, how did RNAi enter your work?
At the time, I was working almost exclusively in mammalian systems and it didn’t seem to operate there, so it didn’t seem to be something with practical utility for me. But, there was a number of labs at Whitehead and at MIT that actually were working on the mechanism, including Dave Bartel’s lab and Phil Sharp’s lab, and through their work I came to learn that … clearly it did happen in mammalian systems. That’s when I became more interested in applying it to my own work.
Can you talk a bit about that application?
It’s become a technology that’s somewhat pervasive in the lab now, and it spans from doing things in mammalian systems using synthetic siRNAs [and] using lentiviral shRNAs to then doing things in Drosophila cells with double-stranded RNA. That’s sort of on a gene-by-gene basis.
Then, in both systems, we’re interest[ed] in doing genome-scale-type of screens with these reagents. So, we’re part of something that’s called the RNAi Consortium — several Boston labs have gotten together to pool resources and obtain funds and knowledge to create genome-wide collections for [mice] and humans of RNAi reagents in retroviral vectors. I’m part of that effort.
I’ve also been working on Drosophila equivalents of that … and making genome-scale collections of reagents for use in Drosophila cells.
How far along are those projects?
In the mammalian one, we have something like 10,000 constructs at this point. I can’t talk too much about it — some of that stuff has not been published. But there are many thousands [of constructs] and plans to make many more.
I should say that making the reagents is really not necessarily the hard part; figuring out how to use them and how to do virus production in a high-throughput way, these kinds of things are more challenging and where we’ll probably put more of our efforts.
The Drosophila side — we have made libraries [and] we have made double-stranded RNA from libraries that are available commercially. We have about half the genome there, but we’re collaborating with people who have larger amounts.
In both systems, we’ve adapted a microarray-based screening method that we call cell-based microarrays, in which we can screen things without wells in a hyper-miniaturized fashion. So, we’re making arrays where we can print these different types of reagents and screen them by growing cells on top of them.
Can you talk about the reverse-transfection technique [used for cell-based microarrays]?
That technique allows you to introduce a nucleic acid into a cell by first printing that nucleic acid on a surface in a certain kind of polymer, and then plating cells on top of it. The cells that land on the little spots where you print the nucleic acid and the polymer take it up, and they take it up in that local place.
Basically, what you’ve done is a transfection with the DNA being on a surface and being introduced locally. So, you don’t need wells to separate individual DNAs. It allows you miniaturize things tremendously and to do, for example, on one standard slide, 5,000 different spots and therefore screen 5,000 different reagents that you introduced to mammalian cells.
What the reagents are can be a lot of different things: They can be over-expression plasmids, they can be plasmids that do under-expression through shRNAs, they can be siRNAs themselves — this is in mammalian systems. In Drosophila, they can be double-stranded RNAs.
[The cell-based microarrays’] utility is that you can do things very quickly because you can print these arrays at many, many thousands of spots and do very quick screens. It’s also very easy to do combination screens where you put one reagent in a solution while you have genome-scale collections on the array. Then, you can look for synthetic effects like synthetic lethals, which are very interesting.
This technology is being commercialized by Akceli, correct?
How did you get involved with the guys there, [such as] David Chao?
It was just through contacts. I meet them and they thought it was something that might be worth working on that could be commercialized. Then there have been a number of people contacting me about … trying to obtain these. As a small lab, it’s hard to make these things available to a lot of people. So, you need some kind of entity like a company that will actually do this.
Do you have a sense of when we might see wide availability of the [microarrays]?
No. I think part of it is that content is still missing — we still don’t have large collections of mammalian reagents. In Drosophila we do and that, as far as I know, is not being commercialized. We’re probably going to have some effort to make them available to academic labs, but in terms of purchasing them, I’m not so sure.
What other areas of your research does RNAi extend into?
We’re using it as a tool for anything we do. Any kinds of studies [where] we want to look at the effect of affecting a gene function, [using] RNAi to knock it down.
Most of what we do is study what’s called growth control — sort of the regulation of mass through a pathway called the TOR [target of rapamycin] pathway. So, we identify new components, or study known components, and we increase their levels, we down-regulate their levels — there’s a number of genes we study in that fashion and we do that both in mammalian cell lines [and] Drosophila cell lines. In all those cases, its possible to use RNAi.
Can you talk a little about the challenges and difficulties facing the technology?
Clearly, in mammalian systems, the fact that a lot of these reagents don’t work [is a challenge]. It’s hard to predict when you make a particular sequence whether it will work or not. We’d like to be able to generate [sequences in mammalian cells] like we can in Drosophila.
In Drosophila, we can basically generate a reagent and we’re almost 100 percent sure that it’s going to work.
In mammalian systems that’s not the case. In mammals you’re not actually using long double-stranded RNA, you’re using these siRNAs and so you basically have many fewer shots on goal. If you put in a long thing into Drosophila, it chops it up into hundreds of these small guys, and there’s a good chance that some of those are going to be good.
In mammalian systems, you’re basically picking one of those. There are some rules [for sequence selection], and they seem to help, but there’re no definitive rules, there’s nothing that you put in and say: Yes, this is the way it goes.
There’s also some talk [about] non-specific effects. I think that can be a problem. We haven’t personally had that issue, but I know other people have.