HANCOCK, N.H. Design automation systems tailored to the task of genetic engineering could prove to be double-edged tools. While they represent a central thrust of the emerging synthetic biology movement, they also can lead to the accidental or deliberate creation of pathogenic biological components.
One expert in the field, Harvard University genetics professor George Church, compared the potential misuse of synthetic biological designs with the danger posed by nuclear weapons. But there is one important difference, in his view it is much harder to build a fusion device than to genetically engineer a pathogen. And the complexity of biological processes also increases the danger of accidents.
By reducing the molecular biology of the cell to a list of standard modules with predictable behavior, professional biodesigners could engineer molecular machines in much the same way that system-on-chip designers create silicon systems. Just as a circuit designer does not need to be an expert in silicon physics and manufacturing processes, the future biodesigner will not need a detailed knowledge of biochemistry to effectively create complex biochemical machines.
"Even if we don't have bioterrorists and teen-age biohackers, we will still create things that do not have the properties that we thought they would," Church said. The problem is that the body has not evolved a general ability to fend off artificial biological agents. "Even if you are genetically resistant and even if you are recently immunized you will have problems with this type of bug."
Church chaired a panel on the problems and opportunities of DNA synthesis at the recent Synthetic Biology 1.0 conference, held at the Massachusetts Institute of Technology earlier this month. A critical question for researchers and entrepreneurs entering the new field is what form technology regulation should take. Church suggested that anyone designing systems with synthetic biological components be required to have a license, which would entail passing basic competency tests.
Licensing might head off the possibility of unintended side effects by maintaining a level of competency among the people in the profession, but would do little to prevent deliberate attempts by terrorists or hackers to create pathogens. The continuing problems the Internet is experiencing with computer viruses that are released secretly give some indication of the problems that synthesized self-replicating systems pose. While the barrier to entry for building a computer or network is very high, once built, it becomes a vehicle for much smaller bits of code that someone with only a low level of expertise can release into the system.
Biological synthesis becomes fairly easy once the basic building blocks the oligonucleotides have been built, so the regulation of the whole process could be centered on licensing and tracking them.
"Our experience with computer viruses is that people that do this kind of thing are rather sloppy, they're not very good at covering their tracks," said Tom Knight, who directs MIT's BioBrick wet lab in the Computer Science and Artificial Intelligence Laboratory. "There is an opportunity here because the oligonucleotides contain a lot of information which can be used to track and monitor what is being done with them."
As an experiment, Knight assembled a list of the oligonucleotides used in his lab and asked himself whether he would be able to predict what was being built with them. "It's not a double-blind experiment, because I already know what we are building with them, but I managed to convince myself that it would be easy to determine that even if I didn't know," he said.
The situation is similar to the nuclear industry, where difficult-to-produce fissionable material is closely tracked and stored in secure facilities. Something similar could be done for oligonucleotide production and distribution.
The potential for danger in this new technology really depends on how effective the emerging techniques will be in actually creating viable biomachines like artificial viruses. Conference attendees seemed to assume that the field will proceed along the same time line as the semiconductor industry. The current state of the art for synthetic biology corresponds to the first steps engineers took to put a few gates on a chip, thus kicking off the chip revolution.
One factor that could speed up the process is the experience that has been gained with electronic design systems throughout the 40 years of VLSI advances. Today's digital circuit designers are uniquely positioned to take advantage of the new field's decoupling of design and implementation. An accident of nature makes it possible to describe cellular processes in terms of the familiar AND, OR and NOT logic operations of digital circuits. So if successful, the synthetic biology movement could lower the entry barrier for electrical engineers for a novel nanotechnology arena-one with broad applications in industry and particularly in medicine. Prototype biodesign systems are already emerging.
But the larger unknown is the speed of the implementation phase. First, standard biochemical modules with standard inputs and outputs will have to be defined. Then some automated, highly parallel manufacturing system will have to be designed that can take a description of a system in terms of standard parts and crank out actual biological components.
"My impression is that biology is still in the dignified style of the English countryside," Knight said. "Practitioners go into the lab and if something works, that's great and if it doesn't, they come back the next day and try again. . . . We have an opportunity to take that stately pace and accelerate it a lot," he said. " Some people here have already developed many of the tools to do that. There is a lot of power and danger here, but I would like to think that the advantages that come with the power outweigh the dangers."
Market pressures are already promptting biotech companies to speed up the DNA synthesis process. Another panelist, John Mulligan, a genetic engineer who started Blue Heron Inc. (Bothell, Wash.) based on his own automated DNA synthesis line, discussed some of the basic enabling technologies that are based on microfluidic chips and robotics.
Mulligan pointed out that speed of synthesis does not just involve the problem of how fast a string of amino acids can be assembled. Chemical reactions are prone to errors, and a major barrier that slows down DNA synthesis is the need to correct errors and verify the correctness of a molecule that can have hundreds of millions of base pairs, Mulligan said.
The large pool of expertise that has been gained by the semiconductor industry in its successful bid to crank out chips with hundreds of millions of transistors is of no use to biological synthesis, however. In fact, the self-replicating nature of biological systems is a built-in manufacturing system, although one that is prone to variations in the form of mutations.
The danger in biosynthesized systems stems from the ability of biological processes to easily support self-replication. Some observers believe that self-replication itself should be strictly banned as the only way to fend off the threat of some engineered molecular system running rampant. But such a restriction would take away much of the power of synthetic biology. And with the pace of innovation, it may already be too late to put the self-replicating genie back in the bottle.