Nobel Prize Winners Deserved Their Nobel Prize

It's a good thing I'm not on the Nobel Prize committee. (Not to mention shocking. The invitation must have been lost in the mail.)

Yesterday I responded rather unenthusiastically to news that Mario Capecchi, Martin Evans and Oliver Smithies -- the pioneers of gene knockouts in mice -- had won a Nobel Prize. Are knockout mice *really *useful, I wondered? Hasn't our newfound appreciation of genetic complexity made that whole knock-one-out-and-see-what-happens approach kind of outdated?

Well -- no, it hasn't. And knockout mice are useful. The first scientist to set me straight was Juan Nunez-Iglesias, a PhD student in USC's Computational Biology program. Commented Juan,

Wow. I can hardly believe your reaction. Mine was: "Really? They hadn't got one yet?" Gene knockouts are the single most useful way to determine gene function. As any systems biologist will tell you (myself included), we would be nowhere without knockouts.

I replied,

Could you go into some more detail on how knockouts are valuable for complex conditions? Whole-genome analyses are routinely giving us disease profiles that involve dozens of genes, each of which seems only to have a small impact. How useful is it to knock out a single gene when its function could be determined by the status of many other genes?

To which Juan said,

... all those studies tell you is that each of those genes is involved in the disease. But how they are involved is usually very unclear. This is the case for essentially all genome-wide technologies. Very often, the knockout mouse gives you the most direct insight into what that gene does, and where it does it. Only once we understand the function of all these do we have a clear picture, and can thus intervene with treatments.

And now back to me:

So the step-by-step would look like this: whole-genome association implicates a bunch of genes. We're not sure what they do. So we do knockout mice for each gene, figure out what the genes do in isolation, and then go about determing how they interact with each other?

And back to Juan:

You've got it essentially right, though knockout mice are also useful without the WGA step. But for this discussion, another important point is that it's super-easy to create double- and triple-KO mice by breeding mice with different KOs -- a great way to study interactions.
Plus, when it comes to testing drugs, the KO mice also come into play
-- you can test how mice with similar mutations to their human counterparts react to various drugs, before doing expensive and risky clinical trials.

And that brings me to a few stories I've written (links below) on the attempts of scientists to breed better lab mice.

So -- a big shout of thanks to Juan, and I hope you readers enjoyed this little peek into the process of journalism, which is also called learning. I certainly don't know everything, I've got my own biases -- namely, against anything that smacks of genetic reductionism -- and sometimes these biases take me in directions that are flat-out wrong.
But I try to be honest about my own limitations and open to new ideas.

(Unless, that is, you're saying vanilla ice cream is better than chocolate, or that the new Animal Collective album isn't pure genius, or that those tight-fitting tapered retro-New Wave jeans aren't freaking hideous. Because then you're just wrong.)

And if you don't want to take Juan's word for it, I've got feedback from a couple biology bigwigs in the next post.

See Also:

• Nobel Prize Goes to Gene Knockout Scientists
• Modern Medical Research Rests on a Bunch of Inbred Mice
• Specially Bred Mice May Hold Keys to Personalized Medicine