Synthetic Biology: It's Not What You Learned, But What You Made

With the news yesterday that J. Craig Venter Institute scientists had built the first bacterial genome from the raw chemical components of DNA, we saw a host of science writers step up to contextualize the work and explain its significance. Our own Carl Zimmer did an excellent job scaling down the announcement to fit within the grand narrative of science as discovery in his Dissection column, "Artificial Life? Old News":

"Creating a new living thing will just mean creating a new set of mysteries," he wrote. "To solve them, scientists will have to plod away with a vast number of experiments. Only then will they get a deeper understanding of life."

In Zimmer's column, there is a purpose, a teleology, to the study of biology: "a deeper understanding of life." But for many synthetic biologists, that's not the primary point of their work. Synthetic biology is to biology what electrical engineering is to physics. In the latter case, both fields involve electrons, but they don't necessarily have the same goals and can't be measured with the same yardsticks. Instead of asking, "What have you learned?" or "What do we understand?" we can ask "What have you made?" and "How did you make it?"

When I interviewed Tom Knight, one of the fathers of synthetic biology, about the international Genetically Engineered Machine (iGEM) competition, he encapsulated the difference between biologists and engineers with a joke:

The biologist goes into the laboratory in the morning and she discovers that the system she's looking at is two times as complicated as she thought it was. Great! she says, I get to write a paper. The engineer goes into the lab, gets the same result and says, "Damn. How do I get rid of that?"

One method of reducing complexity is to simply ignore it. The approach is called "black boxing" and it's common in many types of engineering. A black box is a piece of a system that you view merely in terms of what goes in and what comes out. If you drink five beers (x), you know you will get drunk (y). You don't have to know all the complexities of what the ethyl alcohol does to your brain, you just know, if X
then Y.

A perfect example of "black boxing" is the mechanism the yeast uses to stitch the four long strands of DNA that Venter's team created into the completed genome. A biologist would probably want to understand how that works. An engineer would take it at face value and say, "Great. Let's use it." And that's what they did.

Drew Endy, soon-to-be of Stanford, but a colleague of Knight's at MIT and frequent Wired star, explains synthetic biology like this in a YouTube clip: "It's an approach to engineering biology... it's not the particular application, it's the method. Synthetic biology isn't making a specific thing. It's how you make something."

Zimmer asks of the discovery, "What does it teach us about life that we didn't know before?" But another way to look at Venter's paper is on synthetic biology's terms: What did they make and how did they make it? On the first score, we should be impressed. The combination of techniques yielded a bacterial genome from standard-issue DNA strands that you or I could order over the Internet.

But on the second question--how did they make it--others in the field seem less impressed.

Drew Endy again, this time from his Google News comment:

The technologies pioneered in Japan and at the Venter Institute for genome construction are relatively slow and expensive. We still need to develop "one step"
genome construction methods in order to reduce the costs and turn time of genome construction.

Chris Voigt, my main source for our article, provided a beautiful image for why synthetic biologists are impressed but not awed by the new paper:

There's this great computer in the MIT museum. There's this one computer sitting in there and it is the most intricate woven set of wires. It looks like a rug almost but it was hand put together. That represented the last point when one person could sit there with Radio
Shack components and build the best computer in the world...

"That's what you're seeing in this paper," he concluded. In other words, we just witnessed the end of the beginning for biological engineering. From here on out, as Voigt told me, the construction of genetically-engineered machines will require far more physically-engineered machines and tools.

So, Carl Zimmer and I share a lack of overwhelming excitement about this paper, but for different reasons. For him, it's about the science and the lack of new discovery. For me, it's about the engineering and the lack of a scalable process. Handmade artificial life isn't going to form the basis of the next century of synthetic biology. I'm waiting for fast, cheap genome construction. That will be news, even if it teaches us absolutely nothing about life, because it's how we'll go from the biological equivalents of the ENIAC to the Mac.

"At the end of the 1800s... basically, physics had told you everything that there was to know about electronics," Knight said. "What happened subsequently though, is that we had a century of invention which really was in some sense, not science, but engineering... My outlook is that this century will be dominated by the engineering that comes out of biology."

Image: flickr/mknowles

See Also: