Pages

Saturday, April 13, 2013

DNA laser printing heralds new day for genomics research

DNA Why don’t we use DNA for more things? It’s fairly strong for its size, with some redundant error-checking built right in, and its simple code lets us design strands that act in all sorts of innovative ways. We’ve talked about injecting DNA into sufferers of heart disease to travel through the blood to a damaged area, bind, and self-assemble into a controllable bio-stent that would prop open blood vessels. We’ve talked about storing huge amounts of data in microscopic packages of curled up DNA strands.  We’ve talked about using DNA to micro-suture internal lacerations. We’ve talked about creating artificial life forms with custom genomes. We’ve talked and talked, and precious little of it has come to pass.
The major reason for this is that it’s just too difficult to make DNA. We can design all the bio-stents we like, do all the research necessary to target them, and design specific counter-measures to take them out should they start doing things we don’t expect. None of it matters, though, if we can’t make the strand in question. Our abilities in sequence-reading have been improving along an exponential curve since pretty much the day we published the molecule’s structure, and in that time we’ve made our reading about a hundred thousand times more efficient. It used to cost about $0.01 to sequence a base pair; now it costs about $0.0000001. Synthesis, on the other hand, used to cost about $3 per base pair; now it costs roughly $0.50.
This week, though, a talk by Cambrian Genomics at a DARPA-funded event brought widespread attention to a technology that promises to make DNA synthesis thousands of times cheaper, potentially offering the first real price drop the process has ever seen.
Traditionally, most custom genes are made using bacteria for the heavy lifting, inserting our custom genes so they can piggyback on natural bacterial replication. We can use special enzymes called endonucleases and ligases to splice in or cut out specific bits of DNA, or we can buy pre-made plasmids with easy insertion sites. The problem is that this is very slow, and has difficulty producing long, novel strings. Changing single letters in a sequence is difficult with blunt instruments like e. coli and bovine stomach enzymes, and the sort of fine control of DNA we strive for today requires that sort of precision. As a result, a more direct form of DNA synthesis has always been sought, one which would allow us to build a molecule of DNA one base at a time.
DNA-laser-printer-3This new technology is, in its first step, the same as it’s ever been. Using techniques not unlike those used in sequencing, we affix DNA fragments to a glass plate and grow them base by base. The problem with this has always been accuracy, since the error rate for this type of synthesis is too high to reliably make the precise sequences we require. The novel aspect of the Cambrian Genomics approach comes from the fact that they make many thousands of copies of their sequence, ensuring that at least some proportion will have been made with the proper sequence. Each strand is affixed to a microscopic bead, then read to identify which beads hold strands with the sequence they’re shooting for.
And then, lasers.
This is being called DNA Laser Printing, but that’s not a very helpful definition. It would be more accurate to call it DNA Laser Sorting, as the actual construction process is the same as it ever was, and doesn’t involve the lasers at all. Cambrian Genomics brings in lasers only once the plate is covered with many thousands of DNA-carrying beads, and once each of the beads has been sequenced. With so many copies made, some predictable portion will have been made error-free, and an automated laser flits about over the plate and blasts any beads with a desired sequence off of the plate and into a collector. Once the strands have been washed off of their beads, the experiment is complete; you have a collector full of water that holds only your DNA of interest.
There are any number of applications for this technology. The most high-profile is synthetic biology, which involves creating all new life by substituting a genome of our making for an organism’s natural one. Being able to quickly, cheaply, and with high fidelity create many variants on an artificial genome will allow an explosion in such research. This technology would also be just as applicable to synthetic DNA, so-called “XNA,” man-made variants on DNA that improve or adjust some aspect of their chemistry. Some are hardened against body enzymes for easy transfer through the blood, others are predisposed to create strong, irreversible bonds on command. If we need two very specific bases lined up in very specific places, every time, we can’t afford even small changes that might throw off such alignment. Plus, creating genes from nothing means we can work with viral genes for vaccine research without having to grow the whole, dangerous pathogen.
Precision is what we need to use for nano-scale graphene lithography and self-assembling DNA bricks. It’s what we need to design stretches of DNA that can shrug off attacking molecules but bind specifically to, say, surface proteins on cancer cells. It’s what we need to design substitute genes for use in gene therapy. Genomics has spent a long time looking more like naturalism than experimental science, more about careful observation than making and testing new sequences. That looks like it might be about to change.
Now read: The Quest for the $1000 Genome

No comments:

Post a Comment