I mentioned hooking up small molecules to DNA yesterday. A comment to that post prompts me to write about something I've been thinking about for some time: the work of David Liu at Harvard. I have several of his papers in my files, and he's recently published a long review article in Angewantdte Chemie, for those of you with access to the journal (43, 4848, the International Edition, of course.) Turns out that he has an informative website summarizing the work, too.

In short, what he's been doing is trying to get chemical reactions to go in a much different way than chemists usually do. The inside of a reaction flask is a very weird and specialized environment. We have to really bang on things to make them react - high concentrations, special solvents, catalysts, lots of heat. By the standards of living systems, it's the Spanish Inquisition. Meanwhile, cells make all kinds of things happen by keeping the reactants around in very low concentration (or trickily compartmentalized, a factor not to be ignored), and then sticking them together with other reactants inside the active site of an enzyme. The middle of an enzyme is like a reaction flask that's just big enough for the two molecules, and all sorts of unlikely chemistry happens under those conditions, things that you just can't get away with in bulk solutions.

I should declare my biases here: I find this principle tremendously appealing, and I've had a number of idea spasms in this area myself, which have come on like malarial relapses over the last two years. A number of scattered reports of this kind of thing that have shown up over the last few years; I long to join them. Reducing these brainstorms to practice hasn't been easy, but I continue to think that this general area of research has a huge amount of untapped potential for organic chemistry and drug discovery.

Liu has been taking advantage of the ferocious drive that single strands of DNA have to combine with their complementary partners. He and his group have added chemical linkers to the 3' and 5' ends of complementary strands and decorated them with molecules that could react with each other when they're jammed together by the zipping-up of the DNA ladder. This gives you several interesting possibilities by taking advantage of the huge molecular biology infrastructure of manipulating DNA. Foremost of these is, as I mentioned in the last post, the peerless signal amplification of the PCR reaction, which lets you run everything on microscopic scale and turn up the volume later to see what happened.

Liu's group has tried all sorts of variations on this idea, with different reaction types and different linkers at different positions up and down the DNA chain from each other, and results have been very encouraging. A lot of things are going on. They've found a number of different reactions that can take place under DNA-templating conditions, and they're still expanding the list. They act differently, in surprising ways. Sometimes it's the rate of DNA hybridization that determines the reaction course, and sometimes it's the rate of the small-molecule reaction they're trying to encourage. Along the way, they've shown that some reaction sequences that would normally be incompatible in the same flask can be made to happen in an orderly fashion on the DNA templates.

They've also recently reported using these systems to discover new reactions - splitting and recombining the reactants in classic combinatorial chemistry style, but with that microscale advantage that DNA labeling gives you. You could have thousands of reactions going on in amounts of solvent that a chemist like me wouldn't even notice in the bottom of a flask. Some of these reactions will only work under the DNA-template conditions, which is useful on that side of the research, but not so good for making real-world (that is to say, my-world) quantities of compounds. But some of them look like they can make the leap to non-DNA conditions.

That's just a quick overview - for more details, see Liu's site link above. This is a quickly evolving area, and I'm sure that a lot of neat ideas are waiting to be tried (or even thought of in the first place.) I'm a fan. This is something new, and the more completely new approaches we have to do organic chemistry, the better off we are.