Living cells are amazing things. They created the oxygen we breathe and the fossil fuels that power our world. They provided the organic compounds that form the basis of many drugs and materials. They feed us, live in our bodies, and protect us from other cells and viruses. They can self-organize. They can learn.
It’s clear, then, that the potential range of what biological systems could do is enormous. Among the areas that could most obviously benefit from them are health care, chemical and materials production, environmental remediation, and energy. However, most of the systems that would be useful in these areas are unlikely to occur naturally. We probably won’t stumble upon a cell capable of serving as an artificial blood substitute, for example, or one that harnesses sunlight as transportation fuel. These systems must be engineered.
Synthetic biology seeks to build non-natural systems by adding DNA sequences–effectively, little genetic “programs”–to well-studied cells such as E. coli and yeast. This is, at heart, an engineering problem, one that requires both new “software” (new sequences of DNA) and new hardware (the DNA itself–and the methods for putting it into cells). Synthetic biology has thus far dealt principally with the software. But making the DNA that can be put into cells is difficult and expensive; it has been the fundamental impediment to progress.
Today, long sequences of DNA can be synthesized chemically by commercial vendors at a cost of $1 per base (the DNA “letters” A, T, C, and G). Considering that the sequences we design today are on the order of 10,000 bases, and we want to redesign entire four-megabase genomes, the costs quickly become astronomical. We hope the price will drop, but an alternative lies in the automated assembly of standard biological parts. Here, we don’t synthesize each DNA program with base-level precision. We instead begin with a library of “basic part” DNA sequences. A robot joins these sequences into complete genetic programs using a standard assembly reaction. It is analogous to building electronic devices from a box of transistors, capacitors, and resistors rather than building the whole system at once by lithographic methods. The key is making the technique robust, low cost, and highly automated.
Synthetic biology will really take off once it has transformed itself into an information-driven discipline. The key to that transformation is automated synthesis. The potential is clear–we have no shortage of naturally evolved examples that tell us where the technology can go. We just have to figure out how to take it there.
J. Christopher Anderson, a member of the 2007 TR35, is a postdoctoral fellow in the Department of Bioengineering at the University of California, Berkeley.