Q: What kinds of experiments are enabled by high-throughput gene synthesis?
Church: There are two main applications: pathway engineering and genome engineering. There are probably many more people today doing biochemical pathway engineering, but the genome engineering is also very interesting. Pathway engineering tends to be about libraries while genome engineering tends to be about making a fairly specific goal.
For pathways, what you want to do is gather together pieces of DNA from all over the biosphere that have the right components and make libraries. You fairly quickly get up to a large library size and the application is making large amounts of a chemical or a pharmaceutical or a fuel or something like that. You’re screening for the DNA molecule that confers the best type of enzyme and the best level of enzyme to achieve something.
With genome engineering you generally have a more specific goal for your genome and typically the changes are more scattered about the genome. So far, these new genomes aren’t that different from the original genomes, due to uncertainty about proper functioning as you get too far from the original. Combinatorial complexity could be valuable here, too.
Endy: We see a lot of people building DNA to help them discover and invent therapeutic molecules, enzymes so they can make chemicals, and molecules of DNA that can be used to help us understand how natural living systems work. Those are probably the three big uses right now: the scientific market, the drug market, and the manufacturing market.
There are other people who are creating new markets. There are people who take DNA and don’t use it as a genetic molecule for encoding functions inside living systems; instead, they turn it into an instruction molecule. DNA can be programmed to fold up into two-dimensional or three-dimensional patterns outside of a cell, so you can use it to build stuff, e.g., an inorganic computer where the DNA scaffolds are templates where the inorganic metals go to make the computer. In another application, George just demonstrated the use of DNA for archival data storage of books.
Baker: Protein design has huge promise, but is still in its infancy. What that means is that most of the designs we make will either not have the desired function at all or will have relatively low activities. We have to screen a lot before we get good ones.
We’re making brand new things and we’re not very good at it yet. How well we can make things and how fast we can learn is totally a function of how many designs we can make and how many genes we can get. By being able to test 10 or 100 times more designed proteins with high-throughput gene synthesis, the activities of the best ones would perform much better and, second, we would get a lot of feedback on what works and what doesn’t.
Protein design is not a major industry, but if we could make things on a much larger scale, then it really could be. High-throughput gene synthesis is an enabling technology for that.