Fast FAQs: Running interference

What RNA interference brought to genetics— and what’s still on the wish list for tools

Whitehead Fellow Thijn Brummelkamp uses RNA interference (RNAi), a technique that relies on short RNAs tailored to suppress the expression of given genes, to identify genes that play a role in cancer. Here he talks about RNAi past, present and future. Photo: John Soares

How has RNAi affected basic research?
It allows scientists in basically every laboratory to inactivate a gene, in the most cheap and easy way you can imagine, and study gene function in a variety of settings. This is useful in almost all forms of research where people want to study gene function, and especially useful in cultured mammalian cells.

Are you surprised how quickly it spread?
Yes. The mechanism of RNAi has been revealed in a relatively short time by the work of many laboratories, because researchers are so enthusiastic about the potential applications.

They’re enthusiastic because they were so frustrated.

After the human genome was sequenced, people had a good feeling for the number of genes that were present, but they didn’t know what these genes were doing. And to find out gene function in the more conventional ways was very time-consuming, if you wanted to make knockout mice, or unreliable, if you wanted to use other methods.

RNAi came at the exact right moment, as a quick way to study gene function on a very-high-throughput scale.

What do the prospects look like for the RNAi-based drugs now entering clinical trials?
A main problem for making a good drug is that you need to have a good drug target. Only a subset of genes or gene products can function as a good drug target that is amenable to manipulation by pharmaceuticals.

The major potential advantage of RNAi as a therapeutic is that if you want to inactivate or inhibit any gene product, no matter what kind of protein structure it makes, you can target it and inhibit it with RNA interference. In that way it may open up a tremendous class of drug targets that was totally inaccessible before.

Another potential advantage is that you can design therapeutics by looking at the gene sequence. If you get RNAi to work as a drug, you can imagine that you can now use the exact same protocols for other diseases, just by modifying the sequence of the RNAi molecule.

What are the likely weaknesses?
That depends on the disease. For instance, to suppress HIV, you may need to take therapeutics for the rest of your life. For diseases like that, it may be very difficult to consume or to administer high concentrations of RNA molecules during a lifetime. In other cases, you might be able to cure or suppress a disease in a single episode or a short time period.

There’s also a delivery hurdle. Diseases in tissues that are not easily reached by short RNA molecules will be very difficult to target.

What does your lab work on now?
Since I started my PhD, I’ve been interested in various aspects of cancer research. For example, how do cancer cells respond or not respond to therapeutics? How do they decide when to proliferate and when to differentiate, or when to survive and when to die?

I’m also intrigued by what regulates the size of organs, and how this regulation is involved in the development of cancer. I’ve started working in this area with Fernando Camargo, another Whitehead Fellow.

Are any promising new techniques in the works for genetic analysis?
I’m very much interested in improving mammalian cell genetics so that cultured cells can be used for genetics, like yeast, the C. elegans worm and the Drosophila fruit fly are accessible genetic model organisms.

For cancer research, those existing model organisms are of limited use. Many of the genes that regulate cancer growth are not even conserved in those organisms, either because they are not multicellular or because their lifetime is so short and their cell number is much lower.

Using mammalian cell cultures as a genetic model system is complicated because our genes are present in two copies, and we cannot set up genetic crosses. RNA interference gives us some help, but I’d also like to see other genetic approaches or technologies that would allow us to build these model systems.