Biology Off the Shelf
Drew Endy Th'98 is constructing a tool kit for engineering living systems.
By Jon Douglas '92
In the surrealist René Magritte’s 1936 painting, La Clairvoyance, an artist paints a bird while eyeing an egg. The implication, as the title of the work suggests, is that the man is clairvoyant — that he can predict that the egg he’s looking at will turn into a particular type of bird. But there are other ways to explain the painting. Perhaps the man has had experiences that allow him to draw the appropriate bird for the type of egg he sees. Or, if he knows enough about the general relationship between eggs and birds, he can draw the correct bird for any kind of egg that’s on his table.
To Drew Endy, the different ways of understanding the Magritte painting mirror the transformation he underwent while studying biological systems for his doctorate in biochemical engineering at Thayer School.
Endy was attempting to build a computer model that represented the behavior of bacteriophage T7, a virus that infects the bacteria E. coli. His computer simulation worked well to describe how T7 would grow in its natural state. But when he changed the organization of the virus’ genetic material, the model couldn’t predict the behavior of the new mutant virus. Endy was frustrated by his model’s inability to explain what he was seeing in the lab. “This is one of the best-studied biological systems in the world. But, the most interesting predictions we made were always wrong,” he says. It was as if the painter, faced with a new type of egg, suddenly lacked the knowledge to determine which kind of bird to draw.
After a detailed failure analysis, conducted once he left Thayer and moved to Berkeley’s Molecular Sciences Institute, Endy eventually determined two key problems with the system whose principles he was trying to abstract. First, the functions of only 33 of the 56 genes of T7 were well understood, and second, “the architecture of the parts was screwy,” as if the steering wheel in a car were controlling the radio. The problems led him to the realization that it would be easier to study and model the virus if he built it himself. “We inherit beautiful living systems from nature. But natural living systems are not designed to be easy to understand and interact with. Engineered living systems could be designed that are easier to describe, model, modify, and predict,” concluded Endy. “The question is whether or not engineered living systems will work.”
Since joining MIT’s biology and biological engineering departments in 2002, Endy has been leading the new field of synthetic biology to discover what is actually possible. In a key innovation, he has helped to establish an inventory of 1,205 “BioBricks” — snippets of DNA that can be used as standard biological parts. (The term BioBricks, coined by MIT colleague Tom Knight, is meant to evoke Legos.) To build new living systems, a biological engineer can deploy these components in various combinations, just as a mechanical engineer can pick up screws and bolts at the hardware store before constructing a new automobile engine. Funded by grants, BioBricks are available to researchers free of charge.
“Instead of discovering these proteins in the wild, you can go online and get one off the shelf. The whole collection is designed to work together,” says Endy.
ENDY SAYS THAT THE GOAL OF BIOLOGICAL ENGINEERING isn’t to create life from scratch, but to redefine biology as a technology platform for constructing useful biological systems. Suppose, for example, that a liver cell could be implanted with a counter that records how many times the cell divides. Then, once the counter reached 200, the cell could be programmed to kill itself. That could be a powerful tool to study — and perhaps, fight — cancer. Biological engineering could aid the development of nanotechnology, new energy sources, better materials, and cheaper drugs, among other applications.
“Drew is bringing a new manufacturing technology into the conceptual framework of traditional engineering,” says Thayer School professor Lee Lynd.
In an effort to get more people building with BioBricks, Endy has organized an intercollegiate competition for students from MIT and several other schools, and he has taught a synthetic biology seminar during MIT’s January intersession for the last three years. So far the most complicated biological machines anyone has designed are simple genetic circuits such as cellular clocks and bacteria that have been programmed to form patterns or swim in formation. Endy freely admits that many designs run into problems.
Fortunately, that doesn’t faze him one bit. Endy says he cares much less about building any particular application than figuring out a technological platform that will let anyone build anything that the substrate of biology can physically support. To do this, he concentrates most of his effort on distilling the key principles of the nascent field.
“All of our engineering disciplines are so advanced that we take their foundations for granted,” says Endy. “At no point during my engineering education did anyone sit down and teach me the core ideas that make engineering happen.”
According to Endy, synthetic biology can learn at least three lessons from other fields of engineering, although each might need to be adapted to fit the unique nature of biological systems. The first lesson is to standardize parts. Up until about 125 years ago, Endy points out, you couldn’t replace a screw or bolt in a machine without going back to the manufacturer who made it. Now all fields of engineering take standardized parts for granted. With BioBricks Endy is demonstrating that standardization works in biological engineering, too.
The second lesson is to decouple system design from fabrication. In civil engineering, for example, the architect of a structure is different from the contractor. The same principle can apply to engineered genetic systems: once students design a biological system, they send the information specifying its DNA off to a company for manufacture.
The final lesson is abstraction. For example, when computer engineers want to design a new microprocessor, they don’t start by worrying about the inner workings of individual transistors or where the silicon for the chip is going to come from. Similarly, Endy posits, biological systems engineers don’t need to understand the inner workings of a genetic inverter to make a cellular clock. Instead, they can simply take the inverter devices off the shelf and hook them together.
IF ANYBODY CAN PULL BIOLOGICAL PARTS off the shelf, however, it’s possible that someone might use the technology to cause harm. Endy says that questions of biological risk are ever-present, and that building a constructive community is the best defense against terrorists or disgruntled researchers who might want to unleash harmful biological systems. “The question is: What more should we do to continue to foster a society of individuals who are developing and applying biological technology responsibly?” he says.
Endy is working to keep issues of ethics and risk front and center. When MIT hosted the first International Conference on Synthetic Biology in 2004, the 300 conferees discussed ethical considerations as well as current research in the field. Endy’s synthetic biology competitions involve building “cool genetic machines,” he says, not super strains of bacteria. And he’s instilling a sense of accountability by requiring students to “sign” their BioBrick works with barcodes — though he’s unsure if barcodes make it harder for the biological systems to function properly.
Endy’s view that a responsible community can prevent misuse of technology grew out of a Thayer School seminar given by emeritus professor Arthur Kantrowitz. Kantrowitz had argued that the concept of sustainable development was inherently pessimistic, because if you talked about running out of oil, then you would. Although at the time Endy didn’t buy Kantrowitz’s argument, he came to realize that optimism is as much a self-fulfilling prophecy as pessimism. “It would be irresponsible to develop a new technology and not think about the consequences of success,” says Endy. “We’ll get as many people as possible informed and working together to minimize any risk.”
Endy embraces the idea that biological engineering can be used to change “the human experience we get by default,” but stresses the need to consider the consequences. For example, he envisions that synthetic biology will one day allow parents to design their children’s genetic code to avoid disease. “I’d be perfectly happy to be able to reprogram myself,” he says, “as long as I could reboot if something didn’t work out right.”
Endy recognizes that such grandiose visions for synthetic biology are mere speculation, not clairvoyance—considering that the most complicated genetic machines yet produced have just a few dozen components. But even as he grapples with the specter of new biological problems — such as unstoppable destructive errors in self-replicating machines — he’s more and more confident that the widespread engineering of biology will eventually work. “I’m the sort of person who wakes up every day rethinking everything. But all of this doesn’t seem as mysterious to me as it used to,” says Endy. “There’s a hint of light, way, way down at the end of the tunnel.”
The Making of a Cellular Clock
As a Ph.D. candidate at Thayer School, Drew Endy dreamed of building a clock in a worm but couldn’t figure out how to do it. Meanwhile, Princeton University graduate student Michael Elowitz, now a biologist and applied physicist at California Institute of Technology, managed to construct a simple cellular clock. It took him two years. “We were all inspired by Michael’s work,” says Endy, “and have since been working to figure out how to engineer genetic clocks and other systems in much less time. Any electrical engineering undergraduate could make a simple electric clock in about five minutes. We want to be able to work just as fast.”
An inverter — now available as one of Endy’s standardized BioBrick parts — is the key to cellular clocks. As the term inverter suggests, what goes in is the opposite of what comes out. When the input signal — in this case, the rate at which RNA polymerase molecules move along DNA — for protein A is high, the output will be low, inhibiting the synthesis of protein B.
The cellular clock consists of three interlocking inverters, such that the production of protein A inhibits the synthesis of protein B, B inhibits C, and C inhibits A. Starting with a high level of protein A will produce a low output of protein B. In turn, this low input signal to B makes the output of protein C high. And finally, this high amount of protein C means that the output signal to A will be low. Running through the loop a second time completes the circuit. If the clock is engineered to produce a fluorescent protein only at a high level of A, it will blink on and off as it oscillates through the cycle.
To learn more, follow Drew Endy’s work.
Categories: Featurescomments powered by Disqus