A notable feature of 21st century molecular biology (so far!) is the emphasis on RNA. I've written before about RNA interference, a hugely popular (and hugely researched) way to silence the expression of proteins in living cells. Wide swaths of academia and industry are now devoted to figuring out all the details of these pathways, key parts of which are built into the cellular machinery. They turn out to regulate gene expression in ways that weren't even thought of before the late 1990s, and I've said for several years now that this field is the most obvious handful of tickets to Stockholm that I've ever seen. (Naturally, there are some worries that the whole field has perhaps been a bit over-promoted. . .)

Shutting off the production of targeted proteins is a wonderful thing, both from the basic research viewpoint and the clinical one. The more control you can have over the process, the better, and RNAi has been extremely promising. But as we're learning more about the system, complications are creeping in. Don't they always. . .

It turns out that the small interfering RNAs that are used, and are supposed to be the most efficacious and the most specific, aren't always what they seem. A disturbing recent study used one targeting luciferase, a firefly protein with no close relatives in the human genome. But applying it to the human-derived HeLa cell line showed effects on over 1800 genes - some of which only showed up at high concentrations, true, but none of these would have shown up at all in the ideal world we might have been living in for a while. There have also been experiments with RNAs that have deliberately made with slight mismatches for their intended target, and some of them work rather too well.

Finally, as I mentioned about a year ago, there are reports that these small RNAs can set off an interferon response, suggesting that the technique can cause cells to respond as if they're under infectious attack. As you'd imagine, this can also complicate the interpretation of an experiment, especially if you're already targeting something that might interact with any of these pathways (and plenty of things do.)

None of these yellow flags are particularly large, but there are several of them now and probably more waiting to be noticed. (A good brief roundup of the situation can be found in the November issue of Trends in Genetics, for those with access.) Perhaps as we learn more we'll find ways to obviate these problems. If there's one thing for sure, it's that we haven't figured out all the tricks that RNA is capable of. But the companies that are racing to get RNAi therapies into the clinic are watching all this a bit nervously, hoping that they're not going to be those fools that you always hear about rushing in.