this into circulation, and it will identify malignant cells within the liver, latch on to them, and release cytotoxic drugs.
X: So take the example of the infection sensor and explain the components. How do you build that from the ground up?
DK: You design the synthetic DNA with a section that codes for the green fluorescent protein. [Other parts of] the gene circuitry do the detection of AHL. The circuit also has components like promoters, and we’ve worked on the best type of promoter to make the device work properly. [A promoter is a part of a gene that tells the cell’s machinery to start making a protein based on instructions from the DNA code.] It’s all connected together, and when you put it into the cell, the cell produces the whole device for you.
But to get there, a lot of optimization goes on. At the moment a lot is done in the labs to see what combinations of gene components and chassis [strains of host bacteria for the synthetic genes] works best.
That gets us into what we call foundries. The big thing we’re working on with Jay Keasling at Berkeley is development of foundries to automate what you’d do in a wet lab manually. Taking all the human steps out of it, that’s the objective—get it all done with laboratory robots. Why are we doing that? You can go through every stage of the design and implementation in parallel, the whole work flow, to run through all the combinations of promoters and different cell types. You can test out every single possibility for that device.
X: I’m trying to think of older-world industrial parallels.
DK: You could think of it conceptually as an automotive plant. The whole thing is automated. Except in synthetic biology, the engineering hasn’t been pinned down to the nth degree as it would be in electronics or automotives. So think of a situation where you want the BMW plant producing various versions of a car, trying out different engines, different gearboxes, all in parallel. Of course that doesn’t happen in the automotive industry because the design decisions have already been done.
X: Because this is biology, will there always be a greater amount of uncertainty than in the design of a car to run in a predictable way?
DK: You’d run the foundry to optimize the last bit. You’d ultimately get to a point like in the Toyota production line; the book The Machine That Changed The World years ago analyzed the whole Toyota approach. Instead of letting the car come off the production line then fixing all the problems, every time there was a problem they stopped the production line to fix it. It will be like that.
It gets me into another analogy. A lot of traditional biologists say, “How can you do all this stuff, we don’t understand the biology to the nth degree.” My riposte to that is usually to say, “Well, in 1903 the Wright Brothers flew the first powered flight at Kitty Hawk, NC, and flew for 120 yards.” The whole aircraft industry developed from that. We all fly around the world in all these planes, yet three minutes’ walk from here, colleagues of mine in the aeronautics department are world experts on wing turbulence—which they still don’t understand. It doesn’t stop the aircraft industry from developing. It should be true of synthetic biology. We’re at the early stages but we’re still able to produce stuff.
